Event ·Multi-source

Ramez Naam 分析：AI 能力提升成本两年翻 8 倍

Ramez Naam 基于 METR 报告和内部模型数据估算，Anthropic 每十亿美元投入换来的能力提升随时间呈指数级下降。提升 AI 能力的成本已是 2 年前的 8 倍，且在对数尺度上仍在恶化。Gary Marcus 认为这一结果说明深度学习的改进正以越来越高的代价换取，财务上难以持续。

Current read

Gary Marcus 转发了 Ramez Naam 的成本分析：Anthropic 提升模型能力的花费两年涨到 8 倍，数据来自 METR 报告，争议不小但值得看看算法。

11 sources38° AI heatLast updated 2026/11/9 02:50:08

Evidence

11 sources

Primary sourceGary Marcus

2026/11/9 02:50:08View source

Related source 1theverge

2026/11/7 12:30:00View source

Related source 2Genk

2026/11/7 15:00:00View source

Related source 3Gorden Sun

2026/11/8 04:00:33View source

Related source 4kimmonismus

2026/11/8 10:08:30View source

Related source 5Latent Space (swyx)

2026/11/7 03:04:53View source

Related source 6Latent.Space

2026/11/7 03:16:36View source

Related source 7swyx (AI Engineer)

2026/11/7 03:18:02View source

Related source 8Blognone

2026/11/7 03:23:41View source

Related source 9IT之家

2026/11/7 05:17:43View source

Related source 10The Information

2026/11/7 16:30:04View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

Open the story