AI 战略简报 · 2026-09-24

↓ 下拉刷新

🔥 Grok 4.5 发布

1 条最高 7.6

@amanrsanger · 2026-07-09 02:01

7.6

对这个模型非常期待。相比 composer 2.5，这是一个巨大的进步，而且完全从头训练。能与这个团队合作是一种荣幸...

📘 AI 产品化

2 条最高 7.6

@amanrsanger · 2026-07-28 14:04

7.6

过去一年我们在印度的用户基数翻了三倍，且每位用户的 Agent 请求量高于其他任何国家。Cursor Start 提供了广泛的访问权限...

@amanrsanger · 2026-08-12 01:31

7.4

每一波 AI 产品浪潮都孕育了那个时代的绝对主导者。先是 OpenAI 与对话，接着是 Anthropic 与代码。欢迎 SpaceXAI 与通用知识...

📘 模型公司战略

1 条最高 6.4

@amanrsanger · 2026-08-12 23:52

6.4

这个表现出色的 1.5T 模型！更大且更好的模型指日可待。

尚未被主编选入事件/主题的原始推文（共 21 条，按时间倒序）

@amanrsanger08-29 11:28

RT Tom Brown: Cursor has been a trusted partner of Anthropic since Sonnet 3.5. We’ll continue to increase compute to support Claude models in Cursor ...

@amanrsanger08-29 10:52

RT Michael Truell: We’re sorry to see that OpenAI put out a note saying they plan to block Cursor users from accessing OpenAI models in three months....

@amanrsanger07-09 01:57

RT SpaceXAI: Announcing Grok 4.5, our first model trained specifically for coding and agents. It was trained with Cursor and offers frontier intellige...

@amanrsanger06-16 23:34

Excited to train some very strong models!

@amanrsanger03-25 07:59

It's a good modelCursor: We're releasing a technical report describing how Composer 2 was trained.

@amanrsanger03-21 12:18

We're increasing capacity for Composer 2, 2x more usage all weekend! Try it out!

@amanrsanger03-21 03:41

We've evaluated a lot of base models on perplexity-based evals and Kimi k2.5 proved to be the strongest! After that, we do continued pre-training and ...

@amanrsanger03-21 03:10

RT @Kimi_Moonshot: Congrats to the @cursor_ai team on the launch of Composer 2! We are proud to see Kimi-k2.5 provide the foundation. S…

@amanrsanger03-21 00:57

It's a really good model! Excited for more people to try it

@amanrsanger03-20 02:53

Composer 2 marks the one-year anniversary of our large model training efforts. Since then, we've built an exceptionally talent-dense team of ~40 peopl...

@amanrsanger02-26 03:22

Writing software is becoming effortless. So much has changed and yet there's still so much to go. Cloud agents and demo-based review are a big leap to...

@amanrsanger02-25 03:17

The future of review is not code. It's agents proving their work. For products this is demo videos. And we're slowly discovering what this looks like ...

@amanrsanger02-10 05:19

We trained the best coding model in the world under 1T parameters. Excited for people to try it!

@amanrsanger01-24 11:33

Very excited about this. Rebuilding the primitives of software engineering in an agent-first world!

@amanrsanger01-20 07:04

When does synchronous coding hit diminishing returns to intelligence? I believe we're a few months away from a ChatGPT intelligence bar where >95% of ...

@amanrsanger01-08 03:32

By flushing chat transcripts to disk, Cursor can recall information across a conversation millions of tokens long. A step function better for infinite...

@amanrsanger12-20 02:56

When agents help us write an order of magnitude more code, the bottleneck for software engineering is review.

@amanrsanger11-14 07:20

Hard at work on composer-2!

@amanrsanger11-07 03:06

RT Brie Wolfson: I spent the last 60 days working at Cursor. It's been one of the most thrilling phases of my professional life. There's a lot of myst...

@amanrsanger11-05 03:30

RT Cursor: We've shipped several quality-of-life improvements to Cursor! Details below...

@amanrsanger10-31 06:11

Composer-1 required an enormous amount of science, infra, ML perf, and data/reward work to build. And far more ahead of us! We're growing our small an...