OpenAI最新举措后，“取消ChatGPT”运动声势浩大
"Cancel ChatGPT" movement goes big after OpenAI's latest move
译文简介
在Anthropic公司断然拒绝将Claude AI应用于自主武器和美国公民大规模监控后，OpenAI公司却迅速倒向美国战争部。
正文翻译
There are no virtuous participants in the artificial intelligence race, but if there was, it might've been Anthropic.
人工智能竞赛中并无道德完人，但若真要说有，或许Anthropic公司曾是最接近的那一个。
Large language model tech is built on mountains of stolen data. The entire summation of decades of the open internet was downloaded and converted by billionaires into tech that threatens to destroy billions of jobs, end the global economy, and potentially the human race. But hey, at least in the short term, shareholders (might) make a stack of cash.
大型语言模型技术建立在海量盗取数据之上。数十年开放互联网的全部成果被亿万富翁们下载转化，催生出可能摧毁数十亿工作岗位、终结全球经济乃至威胁人类生存的技术。但话说回来，至少在短期内，股东们（或许）能赚得盆满钵满。
There are no moral leaders in this space, sadly. But at the very least, Anthropic of Claude fame took a strong stand this week against the United States government, to the ire of the Trump administration.
可悲的是，这个领域确实缺乏道德领袖。但至少，以Claude闻名的Anthropic公司本周对美国政府的强硬立场激怒了特朗普政府。
Anthropic was designated a supply chain risk this week, and summarily and forcibly banned from use in U.S. governmental agencies. Why? Anthropic said in a blog post it revolved around their two major red lines — no Claude AI for use in autonomous weapons, or mass surveillance of United States citizens.
本周，Anthropic公司被认定为供应链风险方，并遭美国政府部门强制全面禁用。原因何在？Anthropic公司在其博客声明中解释道，这源于他们坚守的两大红线——绝不允许Claude AI被用于自主武器系统，也绝不参与对美国公民的大规模监控。
It's not unexpected that mainstream governments of any stripe would be salivating at the thought of turbo-charged AI mass surveillance, but it is unexpected that a big tech corp like Anthropic would be willing to take such a strong stance against it in an era increasingly devoid of administrative morality. But hey, there's always someone willing to race to the metaphorical moral abyss in the name of money.
任何主流政府对AI增强型大规模监控垂涎三尺倒并不令人意外，但令人意外的是，在行政道德日益缺失的时代，像Anthropic这样的大型科技公司竟愿意对此采取如此强硬的立场。不过话说回来，总有人愿意以金钱之名冲向道德深渊。
OpenAI CEO Sam Altman and part time supervillain thankfully stepped in to bail out the U.S. Department of War, pledging ChatGPT and other OpenAI technologies to the cause.
幸好OpenAI公司首席执行官萨姆-阿尔特曼——这位兼职超级反派——及时出手为美国战争部救场，承诺将ChatGPT及其他OpenAI公司技术贡献给这项事业。
In a post on X, Altman claimed that OpenAI's models would not be used for mass surveillance, but that claim was immediately contradicted by a U.S. government official, who said that OpenAI's models would be used for "all lawful means." Mass surveillance of American citizens is lawful in "some scenarios" as part of the post-9/11 U.S. Patriot Act, which permits mass harvesting of communications meta data, even if some aspects of it have been curtailed in recent years.
阿尔特曼在X平台上发文声称，OpenAI公司的模型不会被用于大规模监控，但这一说法立即遭到美国政府官员的反驳。该官员表示，OpenAI公司的模型将被用于“所有合法途径”。根据9/11事件后颁布的《美国爱国者法案》，在某些情境下对美国公民进行大规模监控是合法的，该法案允许大规模收集通信元数据，尽管近年来其中部分条款已受到限制。
Anthropic wanted control over the way its technologies would be used, as opposed to relying on the interpretation of laws and legal frx works that even now have been the subject of debate and lawsuits. Altman by comparison is happy to let the U.S. government decide how OpenAI's systems are deployed, which under certain segments of the Patriot Act could quite easily lead to the mass surveillance of U.S. citizens, directly or incidentally as part of provisions on surveiling foreign citizens (which, by the way, is completely legal under U.S. law.)
Anthropic公司希望掌控其技术的使用方式，而非依赖至今仍存争议且涉诉的法律框架解释。相比之下，阿尔特曼则乐于让美国政府决定OpenAI公司系统的部署方式——根据《美国爱国者法案》特定条款，这很可能直接或间接导致对美国公民的大规模监控（此类监控作为监视外国公民条款的组成部分，在美国法律中完全合法）。
The move has sparked immediate backlash on ChatGPT and OpenAI communities online, across threads with thousands of upvotes on reddit of users claiming to be unsubscribing.
此举立即在ChatGPT及OpenAI公司网络社区引发强烈反弹，Reddit论坛上涌现数千点赞的讨论串，用户纷纷表示将取消订阅服务。
ChatGPT recently closed a funding round valuing the company at a frankly absurd $730 billion, with backers including Amazon, Softbank, and NVIDIA. Microsoft professed that it will continue to work with OpenAI, despite saying in an FT interview recently that it would begin building and deploying its own models.
ChatGPT近期完成的融资轮估值已达荒谬的7300亿美元，投资方包括亚马逊、软银和英伟达。微软虽在近期《金融时报》采访中表示将开始构建并部署自有模型，但仍宣称将继续与OpenAI公司保持合作。
Unfortunately, there aren't many other AI companies willing to take a stance against mass surveillance or autonomous weapons. Google removed an explicit ban on the technology last year from its internal rules. Microsoft is cool with autonomous weapons too, as long as a human pulls the final trigger. Amazon has no prohibitions whatsoever besides vague "responsible use" language, and Meta hasn't been shy about courting Pentagon military contracts either. And we all know Palantir is totally for it.
遗憾的是，愿意公开反对大规模监控或自主武器的AI公司寥寥无几。谷歌去年已从其内部准则中删除了对该技术的明确禁令。微软同样对自主武器持开放态度，只要最终由人类扣动扳机即可。亚马逊除了一些模糊的“负责任使用”条款外没有任何禁止性规定，而Meta公司在争取五角大楼军事合同方面也毫不避讳。至于Palantir（美国大数据分析公司），众所周知它完全支持这类应用。
The genie is out of the bottle, so to speak. ChatGPT is great at textual human mimicry but even the most cutting edge models often fail hilariously at even the most basic child-like logic puzzles.
可以说，魔瓶已然开启。ChatGPT擅长模仿人类的文本表达，但即便是最尖端的模型，在面对最基本的孩童式逻辑谜题时也常常会犯下令人啼笑皆非的错误。
Are you looking forward to a world where these hallucination-prone, easily-manipulated artificial intelligence models might eventually decide whether or not you're a threat to national security?
你是否期待这样一个世界：这些容易产生幻觉、易被操控的人工智能模型，最终可能决定你是否构成国家安全威胁？
As long as Sam Altman and his buddies can stay rich, they don't seem to give much of a fuck about it — or you.
只要萨姆-阿尔特曼和他的伙伴们能保持富有，他们似乎对此——或者说对你——根本不当回事。
Since writing this, OpenAI and Sam Altman have been on a damage control mission.
自本文撰写以来，OpenAI公司与萨姆-阿尔特曼已展开危机公关行动。
In an "AMA" style Q&A session on X, Sam Altman claimed that the United States "Department of War" would respect OpenAI's stated "red lines" for not using AI tech for autonomous weapons or mass surveillance of United States citizens, although remained largely vague about how these safeguards would be implemented and maintained.
在X平台的一场“问我任何事”式问答中，萨姆-阿尔特曼声称美国“战争部”将尊重OpenAI公司声明的“红线”——即不将人工智能技术用于自主武器或对美国公民的大规模监控，但对于这些保障措施将如何实施和维护，其表述仍基本含糊不清。
He suggested that existing U.S. law protects against these situations by default, although legal experts have warned that surveillance of non-U.S. citizens permits the collection of data on U.S. citizens in an indirect or incidental way.
他暗示现行美国法律对这些情况提供一定限制/约束，但法律专家警告称，针对非美国公民的监控允许以间接或附带方式收集美国公民数据。
People aren't exactly buying it. It makes little sense for the Trump administration to come out so strongly against Anthropic's stated position, while leaping head first into supporting OpenAI's. The core contention seems to be that OpenAI is happy to let the U.S. Department of War interpret what constitutes "legal," while Anthropic wants to maintain full control over how its technology is used.
公众对此并不买账。特朗普政府如此强烈反对Anthropic公司的立场，却一头扎进对OpenAI公司的支持，这显得颇为矛盾。核心争议似乎在于：OpenAI公司乐于让美国战争部来界定何为“合法”，而Anthropic公司则希望完全掌控其技术的使用方式。
It seems as though Altman is relying purely on hopes and prayers that its technology won't be used for nefarious means — which seems either naïve at best, and dishonest at worst. The current U.S. administration has shown willing to at the very least stretch definitions and precedents outlined in the U.S. constitution and across historical landmark legal rulings. I'm not sure why there's any reason to expect OpenAI's tech wouldn't be co-opted under the guise of "national security," an abuse of power that governmental institutions of all stripes have abused in the past and present.
阿尔特曼似乎纯粹寄希望于其技术不会被用于邪恶目的——这往好里说是天真，往坏里说就是不诚实。当前美国政府已表现出至少愿意曲解美国宪法和历史性法律判例所界定的定义和先例。我不确定为何有人会认为OpenAI公司的技术不会在“国家安全”的幌子下被挪用，这种权力滥用是各类政府机构过去和现在一贯的伎俩。
Since this article, Anthropic's Claude AI app has claimed the #1 top spot over ChatGPT on both Android and iOS. Claude AI is also available for Windows 11.
自本文撰写以来，Anthropic公司的Claude AI应用已在安卓和iOS平台超越ChatGPT登上榜单首位。Claude AI同时也可在Windows 11系统上使用。
人工智能竞赛中并无道德完人，但若真要说有，或许Anthropic公司曾是最接近的那一个。
Large language model tech is built on mountains of stolen data. The entire summation of decades of the open internet was downloaded and converted by billionaires into tech that threatens to destroy billions of jobs, end the global economy, and potentially the human race. But hey, at least in the short term, shareholders (might) make a stack of cash.
大型语言模型技术建立在海量盗取数据之上。数十年开放互联网的全部成果被亿万富翁们下载转化，催生出可能摧毁数十亿工作岗位、终结全球经济乃至威胁人类生存的技术。但话说回来，至少在短期内，股东们（或许）能赚得盆满钵满。
There are no moral leaders in this space, sadly. But at the very least, Anthropic of Claude fame took a strong stand this week against the United States government, to the ire of the Trump administration.
可悲的是，这个领域确实缺乏道德领袖。但至少，以Claude闻名的Anthropic公司本周对美国政府的强硬立场激怒了特朗普政府。
Anthropic was designated a supply chain risk this week, and summarily and forcibly banned from use in U.S. governmental agencies. Why? Anthropic said in a blog post it revolved around their two major red lines — no Claude AI for use in autonomous weapons, or mass surveillance of United States citizens.
本周，Anthropic公司被认定为供应链风险方，并遭美国政府部门强制全面禁用。原因何在？Anthropic公司在其博客声明中解释道，这源于他们坚守的两大红线——绝不允许Claude AI被用于自主武器系统，也绝不参与对美国公民的大规模监控。
It's not unexpected that mainstream governments of any stripe would be salivating at the thought of turbo-charged AI mass surveillance, but it is unexpected that a big tech corp like Anthropic would be willing to take such a strong stance against it in an era increasingly devoid of administrative morality. But hey, there's always someone willing to race to the metaphorical moral abyss in the name of money.
任何主流政府对AI增强型大规模监控垂涎三尺倒并不令人意外，但令人意外的是，在行政道德日益缺失的时代，像Anthropic这样的大型科技公司竟愿意对此采取如此强硬的立场。不过话说回来，总有人愿意以金钱之名冲向道德深渊。
OpenAI CEO Sam Altman and part time supervillain thankfully stepped in to bail out the U.S. Department of War, pledging ChatGPT and other OpenAI technologies to the cause.
幸好OpenAI公司首席执行官萨姆-阿尔特曼——这位兼职超级反派——及时出手为美国战争部救场，承诺将ChatGPT及其他OpenAI公司技术贡献给这项事业。
In a post on X, Altman claimed that OpenAI's models would not be used for mass surveillance, but that claim was immediately contradicted by a U.S. government official, who said that OpenAI's models would be used for "all lawful means." Mass surveillance of American citizens is lawful in "some scenarios" as part of the post-9/11 U.S. Patriot Act, which permits mass harvesting of communications meta data, even if some aspects of it have been curtailed in recent years.
阿尔特曼在X平台上发文声称，OpenAI公司的模型不会被用于大规模监控，但这一说法立即遭到美国政府官员的反驳。该官员表示，OpenAI公司的模型将被用于“所有合法途径”。根据9/11事件后颁布的《美国爱国者法案》，在某些情境下对美国公民进行大规模监控是合法的，该法案允许大规模收集通信元数据，尽管近年来其中部分条款已受到限制。
Anthropic wanted control over the way its technologies would be used, as opposed to relying on the interpretation of laws and legal frx works that even now have been the subject of debate and lawsuits. Altman by comparison is happy to let the U.S. government decide how OpenAI's systems are deployed, which under certain segments of the Patriot Act could quite easily lead to the mass surveillance of U.S. citizens, directly or incidentally as part of provisions on surveiling foreign citizens (which, by the way, is completely legal under U.S. law.)
Anthropic公司希望掌控其技术的使用方式，而非依赖至今仍存争议且涉诉的法律框架解释。相比之下，阿尔特曼则乐于让美国政府决定OpenAI公司系统的部署方式——根据《美国爱国者法案》特定条款，这很可能直接或间接导致对美国公民的大规模监控（此类监控作为监视外国公民条款的组成部分，在美国法律中完全合法）。
The move has sparked immediate backlash on ChatGPT and OpenAI communities online, across threads with thousands of upvotes on reddit of users claiming to be unsubscribing.
此举立即在ChatGPT及OpenAI公司网络社区引发强烈反弹，Reddit论坛上涌现数千点赞的讨论串，用户纷纷表示将取消订阅服务。
ChatGPT recently closed a funding round valuing the company at a frankly absurd $730 billion, with backers including Amazon, Softbank, and NVIDIA. Microsoft professed that it will continue to work with OpenAI, despite saying in an FT interview recently that it would begin building and deploying its own models.
ChatGPT近期完成的融资轮估值已达荒谬的7300亿美元，投资方包括亚马逊、软银和英伟达。微软虽在近期《金融时报》采访中表示将开始构建并部署自有模型，但仍宣称将继续与OpenAI公司保持合作。
Unfortunately, there aren't many other AI companies willing to take a stance against mass surveillance or autonomous weapons. Google removed an explicit ban on the technology last year from its internal rules. Microsoft is cool with autonomous weapons too, as long as a human pulls the final trigger. Amazon has no prohibitions whatsoever besides vague "responsible use" language, and Meta hasn't been shy about courting Pentagon military contracts either. And we all know Palantir is totally for it.
遗憾的是，愿意公开反对大规模监控或自主武器的AI公司寥寥无几。谷歌去年已从其内部准则中删除了对该技术的明确禁令。微软同样对自主武器持开放态度，只要最终由人类扣动扳机即可。亚马逊除了一些模糊的“负责任使用”条款外没有任何禁止性规定，而Meta公司在争取五角大楼军事合同方面也毫不避讳。至于Palantir（美国大数据分析公司），众所周知它完全支持这类应用。
The genie is out of the bottle, so to speak. ChatGPT is great at textual human mimicry but even the most cutting edge models often fail hilariously at even the most basic child-like logic puzzles.
可以说，魔瓶已然开启。ChatGPT擅长模仿人类的文本表达，但即便是最尖端的模型，在面对最基本的孩童式逻辑谜题时也常常会犯下令人啼笑皆非的错误。
Are you looking forward to a world where these hallucination-prone, easily-manipulated artificial intelligence models might eventually decide whether or not you're a threat to national security?
你是否期待这样一个世界：这些容易产生幻觉、易被操控的人工智能模型，最终可能决定你是否构成国家安全威胁？
As long as Sam Altman and his buddies can stay rich, they don't seem to give much of a fuck about it — or you.
只要萨姆-阿尔特曼和他的伙伴们能保持富有，他们似乎对此——或者说对你——根本不当回事。
Since writing this, OpenAI and Sam Altman have been on a damage control mission.
自本文撰写以来，OpenAI公司与萨姆-阿尔特曼已展开危机公关行动。
In an "AMA" style Q&A session on X, Sam Altman claimed that the United States "Department of War" would respect OpenAI's stated "red lines" for not using AI tech for autonomous weapons or mass surveillance of United States citizens, although remained largely vague about how these safeguards would be implemented and maintained.
在X平台的一场“问我任何事”式问答中，萨姆-阿尔特曼声称美国“战争部”将尊重OpenAI公司声明的“红线”——即不将人工智能技术用于自主武器或对美国公民的大规模监控，但对于这些保障措施将如何实施和维护，其表述仍基本含糊不清。
He suggested that existing U.S. law protects against these situations by default, although legal experts have warned that surveillance of non-U.S. citizens permits the collection of data on U.S. citizens in an indirect or incidental way.
他暗示现行美国法律对这些情况提供一定限制/约束，但法律专家警告称，针对非美国公民的监控允许以间接或附带方式收集美国公民数据。
People aren't exactly buying it. It makes little sense for the Trump administration to come out so strongly against Anthropic's stated position, while leaping head first into supporting OpenAI's. The core contention seems to be that OpenAI is happy to let the U.S. Department of War interpret what constitutes "legal," while Anthropic wants to maintain full control over how its technology is used.
公众对此并不买账。特朗普政府如此强烈反对Anthropic公司的立场，却一头扎进对OpenAI公司的支持，这显得颇为矛盾。核心争议似乎在于：OpenAI公司乐于让美国战争部来界定何为“合法”，而Anthropic公司则希望完全掌控其技术的使用方式。
It seems as though Altman is relying purely on hopes and prayers that its technology won't be used for nefarious means — which seems either naïve at best, and dishonest at worst. The current U.S. administration has shown willing to at the very least stretch definitions and precedents outlined in the U.S. constitution and across historical landmark legal rulings. I'm not sure why there's any reason to expect OpenAI's tech wouldn't be co-opted under the guise of "national security," an abuse of power that governmental institutions of all stripes have abused in the past and present.
阿尔特曼似乎纯粹寄希望于其技术不会被用于邪恶目的——这往好里说是天真，往坏里说就是不诚实。当前美国政府已表现出至少愿意曲解美国宪法和历史性法律判例所界定的定义和先例。我不确定为何有人会认为OpenAI公司的技术不会在“国家安全”的幌子下被挪用，这种权力滥用是各类政府机构过去和现在一贯的伎俩。
Since this article, Anthropic's Claude AI app has claimed the #1 top spot over ChatGPT on both Android and iOS. Claude AI is also available for Windows 11.
自本文撰写以来，Anthropic公司的Claude AI应用已在安卓和iOS平台超越ChatGPT登上榜单首位。Claude AI同时也可在Windows 11系统上使用。
评论翻译
很赞 ( 1 )
收藏
Everybody should remember that this is why the free press is important. We'd never know about the deals going on behind the scenes without them.
所有人都该记住，这就是为什么新闻自由如此重要。没有他们，我们永远无法得知那些幕后交易。
@Partridge_Pear_Tree
This administration has solidified certain things for me that are extremely important that I took for granted: The right to vote, Privacy, Autonomy, Education, Press, and Government Disclosures
本届政府让我对一些极其重要、而我过去视为理所当然的事情有了更坚定的认识：投票权、隐私、自主权、教育、新闻自由，以及政府信息透明
We lived with these for so long I just assumed we’d always have it. But now I see how easy it is to lose so much so quickly.
这些东西伴随我们太久，以至于我曾以为它们会永远存在。但现在我明白了，这一切竟能如此轻易、如此迅速地失去。
@_BreakingGood_
It has also made me realize that we cant just rely on old pieces of paper to enforce and maintain these things.
这也让我意识到，我们不能光靠那些老掉牙的文书来推行和维护这些事。
@NapsterKnowHow
And we need a 4th or 5th branch of government to keep the FBI out of the executive. Other branch can be for education, health etc
我们需要设立第四个或第五个政府分支，来防止联邦调查局受行政体系操控。其他分支可以分管教育、医疗等领域。
@4dseeall
writing laws about those things is supposed to be congress' job. but they're complicit to the executive.
制定这些法律本该是国会的职责。但他们却和行政分支沆瀣一气。
@dontyougetsoupedyet
The complicit representatives are the actual, real problem right now. They have had the power to stop all the crimes from day 1, and instead have enabled them. It's one of the few things the system isn't designed to be able to withstand. Because, like, who could conceive of the chances that whole branches of government worth of representatives and hundreds of judges and thousands of lawyers would do this???
那些同流合污的代表们才是眼下真正的问题所在。他们从一开始就有能力阻止所有罪行，却反而助纣为虐。这是现行体制少数几个无法承受的漏洞之一。毕竟，谁能想到整整一个政府分支的代表、数百名法官和成千上万的律师会干出这种事？
@vic20kid
It assumed the representatives would be representing more than themselves
还以为代表们能代表的不只是他们自己呢
@Odd_Soil_8998
Nah you missed the lesson here. You can't rely on the structure of government to prevent tyranny. You have to actually elect mostly good (or at least non-fascist) people in a continuous succession in order to prevent it.
不，你根本没明白这堂课的重点。你不能指望政府架构本身来防止暴政。你必须持续不断地选举出多数是好人（或者至少是非法西斯分子）的官员，才能真正避免它。
@Old-Ad-4897
Don't demand a return to the status quo. Demand more. Minorities have always been the testing grounds for what will eventually hit the majority.
别只要求恢复原状。要争取更多。少数群体向来是试验场，最终波及多数人的事总会先拿他们开刀。
@TaylorMonkey
Don’t forget the second amendment. It’s telling how suddenly exercising it along with your first amendment is “illegal” or “it’s bad I don’t like it” depending on the subjective feeling of the member of the administration.
别忘了第二修正案。有意思的是，当你同时行使第二修正案和第一修正案权利时，政府官员的主观感受就能决定这是“非法”还是“我不喜欢这样，这很糟糕”。
@CrimsonSuede
Jumping in to say this includes local media, too!
我也来插一句，地方媒体同样面临这个问题！
Local newspapers are dying from a lack of subscribers, AI scraping articles, and digital giants hoarding ad revenue. Without local news coverage of key community and local political events, corruption can fester.
地方报纸正因订阅用户流失、AI抓取文章以及数字巨头垄断广告收入而奄奄一息。如果缺乏对重要社区活动和地方政治事件的本地报道，腐败就会滋生蔓延。
My parents have worked for our small rural town’s newspaper for 15+ years. I’ve seen myself how their reporting has stopped local corruption, kept our community informed on political decisions, and brought people together. And I’ve also seen the paper’s decline in that time.
我父母在我们这个乡村小镇的报社工作了超过十五年。我亲眼见证过他们的报道如何制止了地方腐败，让社区居民了解政治决策，并把大家凝聚在一起。同时，我也目睹了这些年报社的衰落。
Like many local papers, they’re on their last legs.
和许多地方报纸一样，他们现在也岌岌可危了。
So where and how you can, please support your local newspaper and media
所以，请在你力所能及的范围内，支持你们本地的报纸和媒体吧
@Partridge_Pear_Tree
Oh I agree! I lived in Fort Worth Texas during the massive power outage (let’s be honest it was a blackout). And the local news did an INCREDIBLE job! They had all the information and made the government accountable. They didn’t hold any punches. I was beyond impressed. They went from reporting feel good stories to the journalism that has been lacking for a while. I have a massive amount of respect for local reporters.
我完全同意！德克萨斯州大停电期间（说实话就是彻底断电）我正好住在沃斯堡。当地新闻台做得太棒了！他们掌握了所有信息，并且让政府负起责任。他们毫不留情地报道，让我印象深刻极了。他们从报道温馨故事转向了久违的真正新闻工作。我对地方记者抱有极大的敬意。
@TuringGoneWild
i get you but we knew about this because Altman tweeted it.
我懂你意思，但我们早就知道了，因为阿尔特曼在推特上发过。
@_BreakingGood_
Altman tweeted the partnership, the free press is what made us learn that their "agreement" of "no mass surveillance or no automated killing" was lip service and non-binding.
阿尔特曼在推特上宣布了这次合作，但正是自由媒体让我们看清，他们所谓“不搞大规模监控或自动化杀戮”的协议不过是空头支票，根本不具备约束力。
@9ersaur
Fastest uninstall of my life
这是我人生中最快的一次卸载
@HiImDan
I let claude do the honors lol
那我就换Claude来代劳吧哈哈
@Thanes_of_Danes
"I believe deeply in the existential importance of using AI to defend the United States and other democracies, and to defeat our autocratic adversaries.
我深信，利用人工智能来捍卫美国及其他民主国家、击败专制对手，具有至关重要的存在意义。
Cool the sloppaganda is working.
好家伙，这宣传套路还真奏效了。
@MealwormMan
Sad you didn’t do it weeks or months ago. The founders of OpenAI have been Trumpers all along
真遗憾你没在几周或几个月前就动手。OpenAI公司的创始人们从头到尾都是特朗普那派的
@omyroj
Y'all have been using that shit?
你们真有人用过那破玩意儿？
@odiephonehome
I’m proud to say I’ve never downloaded the app (not even sure it is an app) or used it ever.
我很自豪地说，我从来没下载过那个应用（甚至不确定它是不是个应用），也压根没用过。
@vee_lan_cleef
It is, it is also a website. I have to say as a Reddit oldtimer it's so strange when people call Reddit an "app", but to a lot of younger people the only interaction they have with the platform is through the app. It started exclusively as a website and is still a website that happens to also have an app. That's my random tangent of the day.
它确实是应用，同时也是网站。作为一个Reddit老用户，我得说听到人们把Reddit叫做“应用”感觉挺怪的，但对很多年轻人来说，他们接触这个平台的唯一方式就是通过应用。它最初只是个网站，现在也依然是个网站，只不过刚好也有个应用版本。以上就是我今天的随机碎碎念。
@88Dubs
Hello fellow "use the website more than the accompanying app"... person
哈喽，同为“用网站比用配套APP还勤”的……朋友
@Mr_strelac
I can understand them in a way... they have problems with profit and they need money.
从某种程度上我能理解他们……他们确实有盈利困境，需要资金周转。
... and on the other hand I can't understand them..this is how they lose their reputation and create problems for themselves, which again when you're not good with money, won't help
但另一方面我又无法苟同……这样搞只会败坏自家口碑，给自己埋雷，到头来财务不景气的时候更是雪上加霜。
@IniNew
If Lockhead, General Dynamics, or any of the other major government contractors are an indicator, with revenues almost in the trillions, this might be the safest bet Scam Altman can make.
如果洛克希德、通用动力这些政府大承包商的情况能作为参考——它们的营收都快接近万亿美元了——那这大概是骗人精阿尔特曼能下的最稳的赌注了。
@Weekest_lixs
They’ll definitely just turn into being dependent on the US Govt. but they could have a been a global company so yeah I guess safe bet, but I don’t know if the investors put in billions for a safe bet. That’s going to get ugly in a hurry on the finance side
他们最终肯定会变成依赖美国政府，但本来是有机会成为全球性企业的。所以嘛，安全牌是打稳了，可我不确定投资人们砸几十亿就为了求个安稳。财务方面很快就要难看了。
@Time_Cupcake_6790
Yeah. None of these corporations care about their reputation so long as they're racking in the profits. Most companies really only use their reputation while it's profitable.
是啊，只要还能大把捞钱，这些企业根本不在乎什么声誉。大多数公司只在有利可图时才会把名声当回事。
Once they have the market cornered, or cash flow locked in, they show their true colors. All of them.
等他们垄断了市场，或者现金流稳了，就全露出真面目了。没一个例外。
@Weekest_lixs
Totally, though in this case it seems openAI is doing it before they fully corned the market.
确实如此，不过这次OpenAI似乎是在完全垄断市场之前就动手了。
@Time_Cupcake_6790
Cause their profits are in the negatives lol they need a cash flow.
因为他们利润是负的啊哈哈，需要现金流。
@bolerobell
It’s a genius move by Sam. OpenAI’s burn rate is much larger than their revenue. They likely wouldn’t have the cash to last another year or two. This virtually guarantees this administration will step in and bail them out as long as they can have autonomous drones that’ll shoot protestors indiscriminately.
山姆这招真是天才啊。OpenAI的烧钱速度远超收入，很可能撑不过一两年了。这基本上等于给政府递了张保命符——只要他们能造出自动扫射抗议者的无人机，这届政府绝对会出手救场。
@UTraxer
HAHhaha "genius", I think you are conflating "genius" with "no-brainer"
哈哈哈哈哈，“天才”，我觉得你把“天才”和“不用动脑子的事”搞混了
Because this literally takes no brains to think up
因为这玩意儿根本不需要动脑子就能想出来
@Sad-Bonus-9327
Safe bet unless the administration changes
除非政府换届，否则稳赚不赔。
@IniNew
I don't know if you've noticed, but Dems don't tend to undo a lot of things that have to do with defense. Obama didn't pull troops out of Iraq. Trump didn't either. It took till Biden, over a decade later to do so.
不知道你们有没有注意到，民主党人通常不太会去动国防相关的事。奥巴马没从伊拉克撤军，特朗普也没撤。一直拖到拜登，过了十几年才办成。
@DragoonDM
I would guess that pulling out of a war is significantly more complex than cutting off AI slop integrations.
我猜，从一场战争中抽身，可比切断那些AI垃圾内容的接入要复杂多了。
@bastardoperator
Trillions? Total revenue for Lockheed was 75 billion in 2025 and 71 billion in 2024. Not even in the ballpark of a trillion and OpenAI just took 100B in funding. Why do people just make shit up to sound cool when they have the answers in front of their face…
上万亿？洛克希德-马丁2025年总营收750亿，2024年才710亿。这连万亿的边都摸不着，而OpenAI刚拿了1000亿融资。答案明明摆在眼前，为什么有些人非要胡编乱造来显得自己很懂……
General Dynamics was 52B last year, so even added together we’re not even getting to 20% of a trillion.
通用动力去年是520亿，就算两家加起来也还不到万亿的20%。
@Ragnarok314159
I am an engineer, make neat things, have a few patents, and currently work in green energy.
我是一名工程师，设计精巧的东西，拥有几项专利，目前从事绿色能源领域。
I could make a lot more designing drone targeting pods and interlocking cluster bombs that blow the legs off children. But I don’t want to do that, rather help build a better world.
我本可以去设计无人机瞄准吊舱和连锁集束炸弹，那种能把小孩腿炸飞的东西，这样能赚得多得多。但我不想那么做，我宁愿帮忙建设一个更美好的世界。
OpenAI and Altman can fuck themselves. You don’t always have to make a choice based on profit.
OpenAI和阿尔特曼都见鬼去吧。人不是非得总围着利润做选择。
@TheGRS
Not an AI engineer but I work in tech, I don't really see how it could turn out well for Altman if the industry starts to give them the cold shoulder. The guard has been changing since last fall, OpenAI is no longer considered best-in-class.
虽然不是AI工程师，但我从事科技行业，如果业界开始冷落阿尔特曼，我真看不出他怎么能有好结果。去年秋天以来格局就在变化，OpenAI已不再被视为业界顶尖。
I was honestly pretty confused by the whole ordeal because I would've figured the government has several initiatives to build their own models. At this point in time it makes more sense to buy, but in a few years I think it will be internalized for security reasons (though they will probably have a hall of a time getting the right AI talent). OpenAI doesn't specialize in government software like Palantir. I guess things could change, but I just don't think going into government contracts is what Altman or anyone at the company really wants to be doing.
说实话整个事件让我很困惑，我以为政府早就启动了好几个自研模型的计划。现阶段采购确实更合理，但再过几年出于安全考虑（虽然他们招揽合适的AI人才恐怕会非常头疼）我认为会转向内部研发。OpenAI不像Palantir那样专攻政府软件领域。或许情况会变，但我总觉得承接政府合同既不是阿尔特曼也不是公司其他人真正想做的事。
@woolybully143
This is so much bigger than money and staying in business…ChatGPT (LLMS as a whole) combined with drones and autonomous vehicles are ushering in the dawn of the robot or drone wars. Historically the biggest barriers to war, was the inevitable death of countless souls of our brothers and sisters, which has always been the least palpable cost of war, one that has often been stood on as an intrinsically good reason to choose peace or diplomacy, and society on whole ha ls agreed to this convention. With the ability to deploy unmanned drones and vehicles equipped with Ai is essentially creating an army who can be controlled by a small group of people, to essentially and perpetually suppress, repress or ultimately destroy civilization.
这早已超越了金钱和商业存续的范畴……ChatGPT（以及整个大型语言模型）与无人机、自动驾驶技术的结合，正在拉开机器人战争或无人机战争的序幕。历史上，战争的最大阻碍始终是无数同胞不可避免的牺牲——这向来是战争中最不被真切感知的代价，也常被奉为选择和平或外交的内在正当理由，整个社会也默认了这项共识。而如今，部署搭载人工智能的无人载具的能力，实质上正在缔造一支可由少数人操控的军队，其将能持续压制、镇压乃至最终摧毁人类文明。
@thegoldinthemountain
This is the one. The unemployment is just the icing on the war industrial cake.
就是它了。失业不过是军工蛋糕上的糖霜。
@phylter99
Sacrificing freedom and dignity for money is a rich guy hobby herein the US. I mean, I get it from that perspective, but that's why we as citizens need to ensure we vote with the money we spend against that.
在美国，为了金钱而牺牲自由和尊严是有钱人的一种消遣。我的意思是，我能理解这种想法，但这也正是我们作为公民需要确保用我们的消费选择来投票反对它的原因。
@Jamesboach
They don't need money. They WANT money. They could go the low and slow method. They could NOT buy the entire supply of Ram. I see countless companies just race as fast as they can to grow and die instead of taking a slow, steady cautious approach.
他们不是缺钱，是贪钱。本可以走稳妥的慢发展路线，也完全可以不抢购全部内存条。我见过无数公司只顾拼命狂奔扩张，最后猝死，就是不肯采用缓慢、稳定、谨慎的经营方式。
@Oscillating_Primate
dexed my account last night
昨晚注销了我的账号
@clownstastegood
dexed mine this morning.
今天早上我把我的删了。
@psychorobotics
Same. Immediately when I saw it, and I never do that.
我也是。一看到就立马这么做了，而我平时从不这样。
@AtariAtari
Never had an account to begin with.
我压根就没开过账号。
@HealthyInstance9182
dexed mine this morning
今早我也删了。
@timify10
I'm done with OpenAI. Corporate greed and pandering to DT will be remarked in history
我跟OpenAI到此为止了。这种企业贪婪和向特朗普献媚的行为必将载入史册。
@Varendolia
On the other hand, Anthropic was brave enough to plant face to them when Anthropic realized they were using their technology to spy on the population and make autonomous weapons, breaching their contract, which prohibited these 2 uses.
另一方面，当Anthropic发现对方利用其技术进行人口监控和制造自主武器——这两项用途都违反了合同禁令时，他们倒是够硬气，直接怼了回去。
The government tried to make them loosen these restrictions, but when Anthropic rejected the idea they started with the threats, government stated that no company working with the government should have ties with Anthropic and they also said they'd use law from 100 years ago to force them to provide those services anyway.
政府试图让他们放松这些限制，但遭到Anthropic拒绝后就开始威胁，声称任何与政府合作的公司都不得与Anthropic有牵连，还扬言要动用百年前的法律条款强行迫使他们提供这些服务。
So after all of that, there's absolutely no fucking way I'd believe OpenAI if they ever say they put any restriction whatsoever on the contracts. And if there's one there's 0 chance they'd care if the other party breaks it.
所以经过这些破事儿，OpenAI要是敢说他们在合同里设了任何限制，我他妈半个字都不会信。就算真有条款，对方违约时他们也百分之百不会在意。