Skip to content
View original post on X: SenseTimeOfficial
Published
22/100

SenseTime's Lin Dahua outlines data, multimodal, and embodied AI priorities

AISenseTime Co-Founder and Chief Scientist Lin Dahua spoke at CITIC Securities's 33rd Investors' Forum on where AI is heading next. He says AI progress depends on data quality, with industry data turned into realistic training scenarios so agents learn faster.
SenseTimeOfficial
@SenseTime_AI
View on X

๐—•๐—ฒ๐˜†๐—ผ๐—ป๐—ฑ ๐—•๐—ถ๐—ด๐—ด๐—ฒ๐—ฟ ๐— ๐—ผ๐—ฑ๐—ฒ๐—น๐˜€: ๐—ช๐—ต๐—ฎ๐˜ ๐——๐—ฟ๐—ถ๐˜ƒ๐—ฒ๐˜€ ๐—”๐—œ ๐—ฃ๐—ฟ๐—ผ๐—ด๐—ฟ๐—ฒ๐˜€๐˜€
AI progress isn't just about scaling up. It's about building smarter.

At CITIC Securities's 33rd Investors' Forum, our Co-Founder and Chief Scientist Dr @lindahua shared his views on where AI is heading next.

๐— ๐—ผ๐—ฟ๐—ฒ ๐˜ƒ๐—ฎ๐—น๐˜‚๐—ฎ๐—ฏ๐—น๐—ฒ ๐—ฑ๐—ฎ๐˜๐—ฎ ๐—บ๐—ฒ๐—ฎ๐—ป๐˜€ ๐—ฏ๐—ฒ๐˜๐˜๐—ฒ๐—ฟ ๐—น๐—ฒ๐—ฎ๐—ฟ๐—ป๐—ถ๐—ป๐—ด.
The value of data lies in its โ€œdensity of intelligenceโ€ โ€” how well it captures the steps involved in complex tasks. By turning industry data into realistic training scenarios, AI agents can learn and improve faster. At SenseTime, over a decade of working closely with customers has helped us continuously improve our models through real-world feedback.

๐— ๐˜‚๐—น๐˜๐—ถ๐—บ๐—ผ๐—ฑ๐—ฎ๐—น๐—ถ๐˜๐˜† ๐—ถ๐˜€ ๐—ฏ๐—ฒ๐—ฐ๐—ผ๐—บ๐—ถ๐—ป๐—ด ๐—ณ๐—ผ๐˜‚๐—ป๐—ฑ๐—ฎ๐˜๐—ถ๐—ผ๐—ป๐—ฎ๐—น.
Describing an image in text can miss subtle but crucial details. That's why we develop native multimodal models, with multimodality built into their architecture and training from the start.

๐—”๐—œ ๐—ถ๐˜€ ๐—บ๐—ผ๐˜ƒ๐—ถ๐—ป๐—ด ๐—ฏ๐—ฒ๐˜†๐—ผ๐—ป๐—ฑ ๐˜๐—ต๐—ฒ ๐˜€๐—ฐ๐—ฟ๐—ฒ๐—ฒ๐—ป.
Much of the real economy operates in the physical world. Combining high-level intelligence with the ability to act in the physical world could accelerate embodied AI, which Dr. Lin expects to gain momentum in the next one to two years. We continue to explore world models that could help AI better understand and predict the physical world.

Ultimately, the true measure of AI progress is how reliably it performs in the real world and the value it creates.

Source details

SenseTime ยท x.comPublished ยท added here

Original title: ๐—•๐—ฒ๐˜†๐—ผ๐—ป๐—ฑ ๐—•๐—ถ๐—ด๐—ด๐—ฒ๐—ฟ ๐— ๐—ผ๐—ฑ๐—ฒ๐—น๐˜€: ๐—ช๐—ต๐—ฎ๐˜ ๐——๐—ฟ๐—ถ๐˜ƒ๐—ฒ๐˜€ ๐—”๐—œ ๐—ฃ๐—ฟ๐—ผ๐—ด๐—ฟ๐—ฒ๐˜€๐˜€