
IT之家 10 月 26 日消息,Anthropic 已向美国证券交易委员会(SEC)提交 S-1 注册文件,为纳斯达克上市做准备。这家 AI 公司早在今年 6 月就秘密递交了上市申请,此次公开的文件披露了更多细节。
据文件所述,Anthropic 今年上半年营收 163 亿美元(IT之家注:现汇率约合 1,088.42 亿元人民币),净亏损达 807 亿美元(现汇率约合 5,388.7 亿元人民币)。《华尔街日报》此前曾依据泄露文件描述过这一情况,其中大部分亏损来自与公司融资票据相关的会计费用。
该文件还列出了 Anthropic 为获取 AI 算力而签订的数十亿美元协议。待所有计划中的算力全部交付后,Anthropic 每月将向 SpaceX 支付约 31 亿美元(现汇率约合 207 亿元人民币),以使用英伟达 GPU;截至 2029 年,相关支出总额最高可达 1,205 亿美元(现汇率约合 8,046.32 亿元人民币)。
值得注意的是,Anthropic 还在文件中写道:“笔者认为 AI 是人类历史上,影响最为深远的工艺之一。如果它得到妥善管理,有望加速科学创新、大幅提高经济增速、让治愈重大疾病成为可能,并应对从前几代人未能攻克的难题。”
Anthropic 的目标是达成上述收益的同时,防范 AI 这项强大工艺带来的风险。AI 相关事故可能造成“灾难性”后果,甚至整个保险行业都可能无法承担此类风险。
本文来源:https://www.ithome.com/1/018/171.htm
若有版权问题,请来信告知:13951431913#139.com 或在右方提交评论。
Thanks to IT Home user 补药吖 for the tip!
IT Home, October 26 — Anthropic has filed its S-1 registration document with the U.S. Securities and Exchange Commission (SEC) in preparation for a listing on the Nasdaq. The AI company had confidentially submitted its IPO application back in June this year, and the publicly filed document discloses more details.
According to the filing, Anthropic reported revenue of $16.3 billion in the first half of this year (IT Home note: approximately RMB 108.842 billion at current exchange rates), with a net loss of $80.7 billion (approximately RMB 538.87 billion at current exchange rates). The Wall Street Journal had previously reported on this situation based on leaked documents, noting that the majority of the losses stemmed from accounting expenses related to the company's financing notes.
The filing also lists multi-billion-dollar agreements Anthropic has signed to secure AI compute capacity. Once all planned compute capacity is fully delivered, Anthropic will pay approximately $3.1 billion per month (approximately RMB 20.7 billion at current exchange rates) to SpaceX to use NVIDIA GPUs; by 2029, total related spending could reach up to $120.5 billion (approximately RMB 804.632 billion at current exchange rates).
Notably, Anthropic also wrote in the filing: "We believe AI is one of the most profoundly impactful technologies in human history. If managed properly, it has the potential to accelerate scientific innovation, significantly boost economic growth rates, make the curing of major diseases possible, and address challenges that previous generations have failed to overcome."
Anthropic's goal is to achieve the aforementioned benefits while mitigating the risks posed by AI, a technology of such power. AI-related incidents could potentially cause "catastrophic" consequences, and even the entire insurance industry may be unable to bear such risks.