中文(ZH) 具身智能还有哪些硬门槛？外滩大会主论坛圆桌回应“降维打击”之问

Embodied AI experts discuss data, models, and commercialization at Inclusion conference

By PulseAugur Editorial · [1 sources] · 2026-09-10 07:15

At the 2026 Inclusion·Wai Tan Conference, experts debated the future of embodied AI, agreeing that its development will not mirror that of language models. Discussions centered on data acquisition, with a consensus forming that a combination of simulated and real-world data, emphasizing physical experience, will be crucial. The panel also addressed the role of large general models, suggesting a collaborative, layered approach where foundational models handle reasoning and task decomposition, while embodied models manage real-time perception and execution. AI

IMPACT Experts predict embodied AI will not follow the path of language models, suggesting a layered approach and a focus on real-world data for commercial viability.

RANK_REASON Panel discussion at a conference about the future of embodied AI.

Read on 雷峰网 (Leiphone) →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

Embodied AI experts discuss data, models, and commercialization at Inclusion conference

How we ranked this

Signal score: 4 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Commentary; Panel discussion at a conference about the future of embodied AI.
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: product, opinion; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

Full methodology in our editorial standards.

COVERAGE [1]

雷峰网 (Leiphone) TIER_1 中文(ZH) · 2026-09-10 07:15

What are the remaining hard barriers for embodied intelligence? The Bund Summit main forum roundtable responds to the question of 'dimensionality reduction attacks'.

过去一年具身智能很火热，却一直没有回答关键问题：训练机器人到底更应该依赖真实数据还是仿真数据？通用大模型会不会“降维打击”具身模型？机器人什么时候才能真正跨过Demo，进入可持续商业化？在2026 Inclusion·外滩大会圆桌论坛《物理智能——分歧与抉择》上，苏度科技联合创始人、CEO韩铮，蚂蚁灵波科技首席科学家沈宇军，自变量机器人创始人、CEO王潜，以及破壳机器人创始人、清华大学交叉信息研究院助理教授许华哲，就数据、智能和场景三场“路线之争”展开讨论。<img…

How we ranked this

COVERAGE [1]

What are the remaining hard barriers for embodied intelligence? The Bund Summit main forum roundtable responds to the question of 'dimensionality reduction attacks'.

RELATED ENTITIES

RELATED TOPICS