The Wisdom Wall

“Yeah, so I think the bar is different. This might be the magic has to get to 10 X. Mm-hmm. Something where 10% is probably within noise and without doubt or something like that, but 10 X is kind of the thing where, like, okay, okay. We, we have to think of things differently now.”

“And maybe I would just add, you know, what happens if you don't have this, and you try to substitute this with an A-plus computer scientist joined with another person who's an A-plus biologist. The problem there is they just do not know what the other one doesn't know, and unless they're like telepathic, this is not…”

“science is something that you can't schedule creativity or the ability to come through a breakthrough. So if you have understood the science of colorectal cancer, what you've learned about a stool test probably is not going to be very useful for a breast cancer test. It'd be a completely different biology. But if…”

“data network effects never go off patent. They just get stronger and stronger and help companies grow even after, even decades after.”

“It's all about, can you make predictions? So if you think you can engineer something, you should be able to make 10 predictions and have three work, not making a 100,000 guesses and seeing what pans out.”

“the intriguing thing is about, ah, data network effect never goes off patent.”

“That traditional biotech is governed by Oom's law, and these are governed much more by Moore's law.”

“if you know, you're just providing services, You may even want to be a therapist as an annuity. Uh, you know, the therapist is there for a long time. Curing someone quickly may not be what you're financially incentive for”

“if you don't let the drug through, we're actually killing people right now by not getting these advances through, and, and I think human beings aren't very good at sort of holding that in their head.”

“One thing that's interesting to think about is that you can think of DNA almost like a one-dimensional image, and so you can use the exact same technology now to put in DNA sequences, and maybe now you're not identifying a face, you're identifying whether someone has cancer from circulating tumor cell DNA.”

“The other thing that happens is that it's very tempting to, once you've built a platform and you actually have an asset that is, let's say, past phase one or phase two, that becomes worth so much. It's a hard financial decision to take money away from that and put it to other things, and then you become a platform…”

“Once you realize that the patient's at the center of it, you realize there's a huge opportunity there. If companies or startups empower the patient, then the patient can drive this whole thing.”

“I think with, especially with machine learning, you could learn features from data without having to share the data itself. And that's useful for IP or for HIPAA and so on. So I think there's a lot of ways that one could contribute to network effects without making your data even publicly known or, or even exchanging…”

“you could get early proof of concept deals, and that will look good. You might have, like, five deals, and it'll be, like, a couple hundred K each. But in reality, those are so easy to get that it's, it's, maybe it sounds more impressive than it is.”

“ironically, I think, you know, most people think about, um, the FDA or, or CLIA being your, your big, ah, concern. I think reimbursement's probably the first place to start, because I wouldn't want to sort of be designing a test without having the confidence that I'll get reimbursed.”

“And when you talk to a biologist, and I think what we're starting to realize is that there's a fundamental conclusion that comes from this, is that biology is so complicated that it's probably beyond what the human being can understand.”

“they've shown that the missing piece is not more hardware, but is software. The missing piece is artificial intelligence and gold standard data sets that can take what the Apple Watch can do, And, and make high quality predictions.”

“And the big, big advance was robots, and so robots really is just not engineering, it's just faster people. You can get to your thousand or 10,000 chances just faster.”

“But for a digital therapeutic like this, you can iterate week after week doing A-B testing, which in the medical world is really a randomized clinical trial that you get to run every week. You make it better and better and better and engineer it and improve it to the point where you can have efficacy that increases and…”

“Um, one of the things that I feel is that people often think regulation is a challenge and in some ways it is, but it's also a huge opportunity. In that if, especially in this new space, you can work with regulators and get there first. You can actually have an opportunity where you can help define what the best…”

“I, I think the reality is that what makes lifestyle and these cultural things, ah, so impactful is behavioral change. And so I think where technology becomes interesting and where we've already seen it become Powerful is to, uh, to create that technology. What makes technology really powerful is its ability to create…”

“I think these so-called social determinants of healthcare that are outside of the traditional healthcare system Is actually a large fraction of what leads to, to mortality and morbidity. And so, AI being able to do behavioral change is such an easy one.”

“But the big win, as I think about it, is like in the sort of value-based care sense, like in the sort of not the sick care fee-for-service sense, but in the truly keeping us healthy sense. If AI can get out in front of things, Then hopefully, you know, we decrease hospitalizations, readmissions, all these things, and…”

“I think it's also not to underestimate the cultural sort of differences between, like, an AI company and a more traditional services company, and getting one to be the other, you know, is, is like getting a boat to be an airplane sometimes, uh, It might be pretty hard, and so it might be something that would be hard…”

“So I think the immediate part is something that is so good, like not 10% better than what you have now, but like 10 X better than what you have now, that the adoption becomes natural. Or so easy to adopt that even 10% better could, could work.”

“Well, so I think the thing that's really underappreciated out of the LLM is like people think of it as like this Oracle or something like that, but I think it's maybe at least for us, I think of it as a UI.”

“where the AI comes in, one idea is a co-pilot, which is like each one of the team members has a co-pilot. But what's interesting about the data is this is like, The, the AI is a peer, you know, contributor, you know, more of the team and has its role that actually everyone feels pretty good about.”

“And I think the general practitioner, sort of, uh, concierge doctor, that tier is kind of a really interesting tier, because largely you're triaging and sending off to specialists. So the AI doesn't have to be an oncologist and a cardiologist and all these things. And so that tier actually alone is kind of really…”

“Like, I'm working on, like, something, you know, in drug design or whatever to make big leaps and bounds, and, um, small things for big cash flows can have a huge impact. So something for clinical trials could be huge, or even just picking, like, the order of rank ordering of clinical trials to Sort of do a better job…”

“if the insurance company is also when providing, they want to keep you healthy. Because healthier people cost less. And of course now we're finally have the incentives aligned.”

“if there's a great drug that, um, ah, that nobody ever gets, you know, nobody ever knows about it. Let's say, let's say there was a cure to cancer, but the FDA didn't approve it. There's no outcry because, ah, no one ever knew about it. But on the other hand, if the FDA let something through that actually, you know,…”

“And so the second part, in addition to the ability for the science to be reasonably well borne out, is that it has to be engineerable, that there has to be enough dials, enough tweaks, enough different things that you can work on to modify such that you can get your 10%, 20% year over year improvement.”

“And this is one of the mistakes I see over and over again with startup founders, is that when they're small, it, it feels like they're all artists. They do a great job of building an org, And then suddenly the org is not what they're comfortable with or familiar with. Because they've built an army and they actually…”

“If you can bring the friction from 30 minutes to a third of a second, that's when the doctor now is actually able to use that information.”

“we've seen companies that are built with science risk, and companies, especially tech companies that are engineering companies, you know, how can we take things from the science curve, which is stochastic and high risk, towards something that's more like engineering, more like, um, grind it out, get it done,…”

“But I think mechanical engineering, electrical engineering, material science, computer science, all these disciplines are pushing into biology. And so instead of steel, it's, it's, it's bone or it's, uh, muscle. But the same principles actually, uh, carry over really nicely.”

“if you're in this engineering curve, the, the, the false positives are actually as important to learning as, as the true positives.”

“And the intriguing thing is that they can do the equivalent of clinical trials, except in computer land, the clinical trial is A-B testing. And they can do this A-B testing, you know, if they wanted to, once a week at scale. They can constantly iterate to make their therapeutic have higher efficacy without having any…”

“In some ways they were right. That was impossible. It was impossible to take existing algorithms and just like shove it down to a very different architecture. You basically had to rethink the problem.”

“in the eighties with the origin of biotech and protein biologics, we already went through, oh, you know, I know what a drug is. A drug's a small molecule. This protein thing, that's weird. I mean, the protein thing, that'll never be a drug. And you can list all these reasons why it'll be a problem. So to some degree…”

“And instead on the, on the digital therapeutic side, you can be running the equivalent RCTs of these randomized clinical trials all the time with super huge N. Uh, but that of course then puts the onus on, on, on those having digital therapeutics to make the case.”

“People don't think enough about is also, is there going to be a huge ROI for your customers? And so, you know, you want to be able to ideally go after, uh, not just the wellness space, but something where you're having a really material impact on the cost of healthcare or the quality of care or ideally both.”

“if you think of this test from a one off point of view, it may look less useful. But if you think of it from a longitudinal point of view or longitudinal in the context of many other tests, suddenly now this is actually potentially useful information.”

“But the beautiful thing about a cloud biology-like setup is that since it's programming on robots, you have the best chance for reproducibility, that rerunning the experiment is rerunning the code.”

“Especially these new modern machine learning methods like deep learning just crave data. And so often you have to reach a critical mass before they can even be used.”

“if you're the one with the big giant corpus, you'll attract the very best data scientists because they'll want to dive into that. They'll come up with the right features and the right ideas, and that will be another sort of effect on top.”

“the simulation doesn't have to be perfect to be provocative. A, there's, you know, the ideal is something where it's perfect, and it gives quantitative predictions, whether of the stock market or of traffic or something like that. But a lot of times things are useful even if it just gives you an idea or a hypothesis or…”

“classification's actually fantastic. If I can say, this is the drug I want to take, or this is the stock I want to invest, um, you know, uh, any of those things are really exciting, and so, uh, I see, sort of, machine learning being, just classification being exciting”

“Whatever the technology is, they got the cool technology, they think this will change drug design, and so then they go to pharma and they try to sell it. And pharma's not convinced yet. Ah, it's like my kids, uh, with new foods, and they don't want to try it until they really know that they really like it. And then…”

“self-insured employers, um, might be a little more motivated because while people may change plans, they change jobs slightly less frequently.”

“So, you know, we've seen companies that are built with science risk, and companies, especially tech companies that are engineering companies, you know, how can we take things from the science curve, which is stochastic and high risk, towards something that's more like engineering, more like, um, grind it out, get it…”

“once you actually have the Legos, Then you get to build stuff. Almost like, um, you know, when people build a bridge, they're not researching steel. You know, they're given the girders, uh, and the, and all the materials, and then they put a bridge together. So I think if people can come up with the parts, people can…”

“I think mechanical engineering, electrical engineering, material science, computer science, all these disciplines are pushing into biology, and so instead of steel, it's, it's, it's bone, or it's, uh, muscle, but the same principles actually, uh, carry over really nicely, and if you look at on the academic side, uh,…”

“But if you're in this engineering curve, the, the, the false positives are actually as important to learning as, as the true positives.”

“if you break it up into little bits, any little bit isn't so bad. And, and can be engineered. And, and, and you sort of do it step by step by step by step. And I think that's, for me, the inspiration for how to take some big crazy thing, uh, like going to the moon, that if you did it from, like, a screening…”

“Evolution has created a ton of features. It's an amazing software engineer of sorts. But, you know, with these features come bugs.”

“And if you think about it, biology is all about technical debt. Biology is trying to get that MVP out. You're trying to, like, survive the T-Rex coming after you. You know, you're not gonna be able to make things perfect.”

“I think you might be seeing where I'm going with this, is that that's a lot like what we have to do in biology. You know, we have to get into the code which is a mess and understand it, and just understanding the code of biology is something extremely difficult, but then once we've done that, that's when the…”

“What I love about software is that software, we can de-risk at every stage of investment, and we can see greater and greater revenue and help the company get bigger and bigger.”

“Similarly, we could toss out the healthcare system and maybe start fresh and it, and that'd be an intriguing sort of, ah, thought discussion. But in reality, we can't do that either. And so what we're gonna see is people working within the system.”

“if I asked you, like, 25 plus 17, that's really easy to do. If I gave you that same problem in Roman numerals, you'd probably have to think about that back into Arabic and then do the computation and put back into Roman. Some representations make computation natural.”

“when you can engineer something, when you can industrialize something, you can make it better, you know, 10% year over year, 20% year over year. And with that type of interest rate, so to speak, Something that starts off kind of crappy, uh, kind of ugly, or just basic can get cheaper exponentially, can get better…”

“there's a learning aspect here of machine learning, which is intriguing, that as you get more data, you get better, and that's something that's really not like any other test, where, you know, a lipid blood test doesn't get better as you have more patients there”

“Well, it turns out that identification is very similar to identifying someone in a picture.”

“typically at first industrial products are worse that the chairs, the tables being made from these factories are not beautiful artisanal bespoke things. They're, you know, kind of crude looking, kind of ugly, uh, but much cheaper”