BestBlogs.dev

机器人终局:NVIDIA Jim Fan 谈 Embodied AI 的下一步 (opens in new tab)

📌 One-Sentence Summary NVIDIA 的 Jim Fan 认为机器人会沿着 LLM 的 scaling 路径前进,通过 world models、第一视角人类视频、neural simulators、World Action Models 和 real-to-sim-to-real 数据管线实现突破。 📝 Summary 这场 Sequoia 对 NVIDIA Jim Fan 的访谈,是一条高价值的机器人与 embodied AI 路线概览。Fan 的核心观点是 Great Parallel:机器人可以借鉴 LLM 的成功路径,从 next-token prediction 转向 next-world-state prediction,再对真实机器人需要的 action slice 做 fine-tuning 和 reinforcement。访谈有价值的地方在于它明确指出了瓶颈和研究方向:当前 VLA 模型过于 language-heavy,对物理和动作理解不足;teleoperation 数据难以 scale;人类第一视角视频和 sensorized da...

Read the original article
Sign in to keep reading the full article.

Keyboard Shortcuts

Navigation

Next / previous post
j/k
Open post
oorEnter
Preview post
v

Post Actions

Love post
a
Like post
l
Dislike post
d
Undo reaction
u
Save / unsave
s

Recommendations

Add interest / feed
Enter
Not interested
x

Go to

Home
gh
Interests
gi
Feeds
gf
Likes
gl
History
gy
Changelog
gc
Settings
gs
Discover
gb
Search
/

General

Show this help
?
Submit feedback
!
Close modal / unfocus
Esc

Press ? anytime to show this help