John Tsang

Where the AI came in is very intriguing.

We thought, well, now let's represent each person just as a long string of numbers.

You now have a map that you can place each person on that map, right?

If you take the very first dimension that explains the most amounts of differences across all of these people that we have looked at, correlated almost perfectly,

with this probability of someone being healthy.

AI was rediscovering, in a way, our probability of being healthy axes.

And that's super intriguing because it suggests that that's the most fundamental wavelength, in a way, in the system.

And once the AI is done, we can go back and ask, what did you use in order to compute this probability?

That's how we figure out how some of the underlying parameters are contributing.

Actually, that's where we're moving next right now.

It's called the Human Immunome Project.

The idea is we want to measure the entire world's immune system to increase representation of individuals with very different genetic backgrounds, very different exposure histories, facing very different kinds of stresses and so on, right?

So that would help us to train these models to better represent all of the individuals on earth in terms of their immune system.

No, that's not part of the plan at all.

Our mission is to make this openly available

for researchers and anyone around the whole world.

And second, we will be working with all the local scientists and jurisdictions about data privacy and everything.

And in the sense that, of course, we want to protect the privacy of the participants.

So the goal is not to take this data and sell the data to any specific entity or use it to benefit only a specific entity.

Imagine a world where you could monitor how things are going in your body and when there are signs that your body is going in the wrong direction.