返回 2026-10-02 编辑版
发现与解读中文优先 · EN 原文待分类

Tavus 发布：Griffin 实时视频对话模型自述有 48% 体验者误认为真人

介绍Griffin，这是第一个通过视频图灵测试的模型。 与它进行实时交流的48%的人认为它是真人。之前的系统通过率不到3%。它在……中排名第一NVIDIA全双工的基准AI视频。 这是第一个人类互动模型（HIM).

查看英文原文

Introducing Griffin, the first model to pass the video Turing test. 48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video. It’s the first Human Interaction Model (HIM).

Introducing Griffin, the first model to pass the video Turing test. 48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video. It’s the first Human Interaction Model (HIM).

事实主体／归属Tavus@tavus
发现来源X @_akhaliqx-akhaliq
打开原始出处
本期解读 · 产品与采用

垂直应用开始用交互、集成和持续服务竞争，但采用证据仍以厂商自述为主

为什么值得看

应用层不再只比较回答质量，而是在真人感交互、行业连续服务和集成覆盖上建立门槛。真实机会需要继续用留存、任务完成率、付费和故障率验证。

最强反方

图灵测试、旅行者数量和集成规模都来自发布方口径；三条信号没有统一样本、客户结构、活跃度、收入或长期效果数据。

EVIDENCE

原始出现与传播对照

2 条出现 · 本次 AIHOT 未精确命中

发现与解读

介绍Griffin，这是第一个通过视频图灵测试的模型。 与它进行实时交流的48%的人认为它是真人。之前的系统通过率不到3%。它在……中排名第一NVIDIA全双工的基准AI视频。 这是第一个人类互动模型（HIM).

介绍Griffin，这是第一个通过视频图灵测试的模型。 与它进行实时交流的48%的人认为它是真人。之前的系统通过率不到3%。它在……中排名第一NVIDIA全双工的基准AI视频。 这是第一个人类互动模型（HIM).

查看英文原文

Introducing Griffin, the first model to pass the video Turing test. 48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video. It’s the first Human Interaction Model (HIM).

Introducing Griffin, the first model to pass the video Turing test. 48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video. It’s the first Human Interaction Model (HIM).

直接发布标题Tavus@tavus打开该层出处 ↗

Introducing Griffin, the first model to pass the video Turing test. 48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video. It’s the first Human Interaction Model (HIM).

来源
X @_akhaliq · x-akhaliq
证据角色
x_project_network
时间状态
known
用户判决
未判
查看该来源原文raw/x-akhaliq.json6bee3d4443f903d8ee625290db8de1acff69c0962d2f247f2eb9e2f423341fd3
发现与解读

介绍Griffin，这是第一个通过视频图灵测试的模型。 与它进行实时交流的48%的人认为它是真人。之前的系统通过率不到3%。它在……中排名第一NVIDIA全双工的基准AI视频。 这是第一个人类互动模型（HIM).

介绍Griffin，这是第一个通过视频图灵测试的模型。 与它进行实时交流的48%的人认为它是真人。之前的系统通过率不到3%。它在……中排名第一NVIDIA全双工的基准AI视频。 这是第一个人类互动模型（HIM).

查看英文原文

Introducing Griffin, the first model to pass the video Turing test. 48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video. It’s the first Human Interaction Model (HIM).

Introducing Griffin, the first model to pass the video Turing test. 48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video. It’s the first Human Interaction Model (HIM).

直接发布标题Tavus@tavus打开该层出处 ↗

Introducing Griffin, the first model to pass the video Turing test. 48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video. It’s the first Human Interaction Model (HIM).

来源
X @ycombinator · x-ycombinator
证据角色
x_project_network
时间状态
known
用户判决
未判
查看该来源原文raw/x-ycombinator.json74365f29584aba9cfe496f63032be2b6676fc756248af0d237df55eb8e5d6419
Cluster IDevt_1330bd371ebbb28a9ebba84e精确去重依据canonical_url事件状态未判