AI Investor Forum: Gus Tai on AI in Healthcare (Part 6)

Sramana Mitra: Some of you here would remember, Gus would definitely remember that many years ago, I came up with a formula for the future of the web. It was called Web 3.0 = (4C + P + VS).

Gus Tai: Yes, I remember that.

Sramana Mitra: The four Cs were context, commerce, community and content. P was personalization and VS was vertical search. So if you take that formula and apply it to different disease conditions or different niches, in this case, menopause stage of life, which is quite a large niche. There are maybe smaller niches.

Recently, one of our relatives got diagnosed with Guillain-Barré syndrome, which I knew nothing about. I’ve never heard about it, but turns out it’s a serious disease. The level of information is very low.

So there could be many, many such niches where you could create a specialized telehealth clinic that deals with that particular disease condition or stage of life. There are various ways of determining context.

Midi health has determined menopause as the context around which to build this new telehealth clinic. Telehealth is a very interesting opportunity, and it is something that you can get to proof of concept or some level of product-market fit, and can then scale with funding. Or if you want, you can work in bootstrapping mode all the way, depending on the size of the niche.

Gus Tai: Yes. I would add that it would become even more economic to do that as AI rolls out. So, AI can be used as a capability to fundamentally transform or offer a completely new type of service.

I think, over the next several years, we’ll just see how it makes the workflow more efficient. I had mentioned that Epic and other companies can ingest a doctor visit and provide metadata. So long as an expert is reviewing the metadata, that summary can have great fidelity and be very inexpensive. You don’t have to worry about the hallucination problem if someone’s reviewing the data.

Sramana Mitra: Absolutely. This is a very important point because in my model of Web 3.0 = (4C + P + VS), the content piece with AI becomes very powerful.

You insert generative AI trained on interesting content, you have to definitely watch for hallucination because in healthcare, hallucination cannot be permitted. This is a mission critical area. It has to be somehow controlled and that could be done. There are different ways that people are dealing with hallucination.

One of them is small language models, but it can be done. Then the content becomes really powerful, interactive content. Generative content becomes really powerful and can be a huge scaling engine in this particular worldview.

The other point to remember from an opportunities to go after is, telehealth is necessary for the global population at large.

The vast majority of global healthcare in every dimension and every context is underserved. So, if you want to do a telehealth clinic around a certain context in Nigeria; that’s an opportunity. You can do one in India; that’s an opportunity. Then you have to layer in language elements on top of that.

Vast majority of the rural populations don’t understand English, so you have to run your clinic in the local vernacular language, and at a price point where those medical systems can operate.

There are lots of nuances to all these problems, but this is an infinitely large startup opportunity. Extremely rich.

Gus Tai: Very much agree. That was well said, Sramana.

This segment is part 6 in the series : AI Investor Forum: Gus Tai on AI in Healthcare
1 2 3 4 5 6 7 8