Skip to playerSkip to main content
  • 8 months ago

Category

🤖
Tech
Transcript
00:00Gong Ying Ying. She's of course the founder of Yidu Tech. Ying, good to have you with us.
00:05Thank you. How do you weigh in on this debate? Is China leading in terms of health tech or is it the U.S.?
00:10Well, I am not comparing those two countries. I can share with you my 2025.
00:19So last year when I was in Douglas, there was the deep sick moment.
00:25And before the deep sick moments, Yidu was actually building our own health care foundational model.
00:32And at the time, we actually thought there is going to take us another three years, billions of R&B investments.
00:40So we can use our own foundational model to create Dr. Copilot.
00:47But luckily, we had our deep sick moments.
00:50And this year, which means so busy deploying Deep Seek, Tianwen, together with our knowledge graph and also our own health care model.
01:02And one use case we have is the doctor's agents.
01:06So we provided the AI agent factory to the hospital.
01:11And then we create agents for the top hospitals, doctors.
01:17And now that, you know, just after a year, some of the hospitals, more than 50% of the doctors would be very, very comfortable to have three to five agents working for them.
01:29Some does tumor staging, some does patient education, some does sorting out their disease data set, you know, doing timelines and sorting out the EMR records for very long NCD diseases patients.
01:43So, you know, it's still very easy here.
01:46I guess the question is, where do you go from here?
01:48Can AI replace doctors?
01:50And how far away are we from that?
01:52I have a point.
01:54I think AI will never replace doctors.
01:59And, you know, for a few reasons.
02:02One is, you know, it's too early to hand over the will to a person that we don't know who is driving the car, right?
02:10There's a lot of, you know, this debate on AI safety and also compliance ethics.
02:17And also we need to train, we need to give opportunities to the younger doctors to grow with the AI.
02:25And so I don't, you know, for that reasons.
02:28And number two is, I think that there's a lot AI can learn from the top physicians and also real life, everyday clinical, you know, studies.
02:39We're seeing the health tech space in China really booming.
02:42What's behind that?
02:43Is it regulation?
02:45I mean, how are regulations versus the U.S.?
02:48So in China, I think there's a, you know, we have been very lucky that we have a very, very big market.
02:56When we have, when you do launch your products and we go to our clients' hospitals in Beijing and they look at our product and they probably don't like it.
03:04But then we take this, the same product and say, okay, maybe we'll sell it to Chongqing.
03:08And then if we get very unlucky that the Chongqing hospital don't like it, okay, we'll go to Shenzhen.
03:14And someone would like them and say, okay, maybe let's modify, let's work together.
03:19And then we change our products and then, you know, and then we're able to sell to the entire country.
03:25But is the regulatory environment supportive?
03:27Yes, very supportive.
03:28The government has launched a lot of policies supporting AI plus and healthcare is one very big sector.
03:37And also we have an aging population problem and also 2030 healthy together policy.
03:44So those policies are encouraging entrepreneurs to do business in this field.
03:51And I know that Yidu Tag was actually founded to address the issues of the aging population targeted for 2030.
03:58What do you think, what illnesses can AI solve in an aging population like China?
04:05So when I started the company in 2014, I actually, before 2014, I did a two years field study in Japan, Israel, the U.S.
04:18And I realized that medical big data and AI technology is the answer to solve the productivity of a healthcare industry.
04:30So if you look at all the sectors that Yidu is focused on, the first one is the data infrastructure for hospitals and also CT regulators.
04:43And then you have a lot of agents for people to work very efficiently on top of that.
04:50And then you have the AI CROs.
04:52And the reason that we started this AI CRO is because we want to solve the problem of 10 years, a billion dollars or multiple billion dollars of cost.
05:01And a lot of the rare diseases drugs are so expensive.
05:04It's because of the way that we develop those drugs can be more efficient.
05:09And also we can apply it to a bigger market.
05:12And then we have the AI insurance solutions that we've launched, the inclusive supplementary insurance in multiple cities.
05:19And in some of the cities, it takes 1.2 seconds to get your claims back.
05:25I'm just wondering whether there are two different approaches to how the U.S. and perhaps China approach AI.
05:33Is it about cost, essentially, for China and for the U.S.?
05:37It is about all that's in pursuit of cutting edge technology.
05:40Is that a fair way of looking at the two ways of building AI?
05:45Well, at least that's not the way I'm looking at it.
05:48And I really think that healthcare is a sector that humans still have understood very little, like less than 5% of all diseases around the world.
06:02And there are so many rare diseases, maybe that just, you know, even if we understand them, it's not commercially viable to develop drugs to cure them.
06:13And so we need a bigger market to have those drugs being developed.
06:19And also, you know, the climate change.
06:22And there are lots of different diseases that we don't understand, right?
06:25So even in those areas, we've launched about 400 different specific disease data sets.
06:34And for those disease data sets, we use our AI to do a lot of research on, you know, rare diseases and diseases that we don't understand.
06:44And the PIs around the world are working for solutions for them.
06:48So a lot of time when you publish those papers and doing those research is, you know, it's not just about the cost.
06:56It's also about, you know, the right thing.
07:00I'm just wondering what you're most excited about over the next 12, 24 months.
07:04What should we be looking out for when it comes to e-do tech?
07:07I think for our company, I think it's just we, you know, within one hospital, you're seeing doctors getting very comfortable with agents.
07:18And then you're seeing multi-agent tasking, agents working with each other.
07:22And then you're seeing, you know, agents working with the medical operating system.
07:27And then that medical operating system is going to embed more data and more intelligence around the world.
07:35So you're really seeing the change of workflow of a health care system and over the next maybe 12 to 36 months.
07:44And also, you know, there will be more and more drugs being developed within our AI CRO.
07:50So those will be exciting moments for me.
Comments

Recommended