Dr. Joe Geraci – Nurosene, NetraMark, Data Science

Transcript:

Im a Mortal Episode 9: Joseph Geraci – Nurosene, NetraMark, Data Science Transcript

Speakers: Joseph Geraci (Guest), Sufal Deb (Host), Marvin Yan (Host)

[MUSIC – Im a Mortal Theme]

Joseph Geraci  0:26  

My name is Dr. Joseph Geraci. I am a mathematician, a medical scientist, a quantum machine learning specialist, and an entrepreneur. Until recently, I was the CEO of NetraMark Corp. It just got acquired by a company called Nurosene, its ticker symbol is MEND. NetraMark, which I’ve been developing over the last five years, is a unique way to understand patient populations for complex disorders. Stuff like oncology, neuropsychiatry, aging, and things like that. Diseases where it’s difficult to label. One of the things that I’ve been focusing on heavily is in developing next-generation machine intelligence that can actually help prepare datasets to feed to machine learning. So we use machine intelligence in a very unique way. We use it to deal with heterogeneity, and understand the patient populations from the patient level in a very precise way. Then, this machine prepares novel data sets and it gets pushed into machine learning. The other thing we focus on is this cool platform we call NetraAI, which allows you to actually interact with complex patient populations. You can literally augment your ability to understand the disease state with a lot of precision. It’s very cool. I’m also a professor of Molecular Medicine at Queen’s University. I’m also a professor at Augusta University in Georgia, USA. That’s me.

Marvin Yan  2:10  

Okay. Wow, a man of many hats, I see. We definitely have a few questions coming up, especially on– I think, last we spoke, Joe, we talked about your Alzheimers paper, we did read that. I do have some questions. Before we get there, given that our podcast is called, Im a Mortal, which is a bit of a play on the word “immortal”. What does the word “immortal” or “immortality” mean to you?

Joseph Geraci  2:30  

What it means to me is the ability to choose when to terminate my life. That’s what it would mean to me to be immortal. Am I done? Instead of the current situation where we age, we break down, and we have little choice. There’s a benefit to not having a choice, which is, I think, you live life more fully and you’re forced to look deeply into the nature of reality. It can drive some very deep spirituality, which is very satisfying. That aside, it would be nice to choose when I die.

Marvin Yan  3:10  

I’m curious, do you think we should all have the choice for when we pass, then?

Joseph Geraci  3:15  

I don’t know. I don’t know. That’s a great question. I think where– the right that we should all have is, we should be able to reprogram ourselves so that a lot of the stuff that takes people down early gets eliminated. Maybe that is something that we can all agree upon. No more ridiculous neurodegeneration, no more cancer, no more of that type of stuff. Age-related disorders, wasting away. I think what we should focus on that aspect is aging really well, which some people do. It’s one of the things I studied. We call them SuperAgers. I think that making everybody a SuperAger, I think that’s the right– there it is.

Sufal Deb  4:08  

Joe, before we jump into everything about AI, your research, your company, and whatever else we want to talk about, I want to ask about your journey and how you got to where you are. In your intro, you mentioned things such as mathematics, oncology, neurodegenerative disease, physics, artificial intelligence, and I’m sure I could go on and on with some of your help. Where exactly did your journey begin? Where did your interest start? 

Joseph Geraci  4:30  

Yeah, so when I was a kid, I was very, very science-oriented. I won some science awards but the real place where it took off was when I begged my parents to buy me a computer from RadioShack when I was 11 and they did. I got this CoCo 2 computer with 16K of RAM or something. It was amazing. I just loved programming this thing, but I wanted to make it intelligent. I wanted it to be able to converse with me and so I tried to develop my own methods to do that when I was a kid. Eventually, I had this realization that the only way for a computer to start to emerge any semblance of intelligence is to give it the scaffolding so that you can leave this thing on, and it can learn from data. I had that realization when I was 12. I remember years later when I was doing machine learning, I’m like, “Hey, this is kind of the stuff I was doing when I was 12. I mean, just a lot less sophisticated. ” I didn’t know how to build the proper models and stuff but I had that paradigm. 

Ever since then, I was just obsessed with mathematics and science. I ended up doing math degrees in my undergrad and in my master’s. Then, during my Ph.D., I was asked to look at a problem in quantum computation and the US government became interested in one of my proposed solutions. The next thing I knew, I moved from Canada to the US and lived in Los Angeles. I was thinking about what are the outer limits that are going to be possible with this next generation of computations, quantum computation. That’s where I got started. Then, once I did that work as a Ph.D. student and graduated, I switched to medicine. I did a postdoc in oncology, another postdoc in computer science, machine learning for medicine, and then another postdoc in neuropsychiatry. That was kind of my journey, I was always just hyper interested in computer science, biology, and medicine. I managed to merge them.

Marvin Yan  6:50  

Oh, well, forgive me for prodding further, but I feel like I want to know a little bit more because I know just because you like k-nearest or decision trees doesn’t mean you necessarily are into longevity or aging. At what point did the concepts of longevity or aging start to go on your radar?

Joseph Geraci  7:07  

Yeah, no, that’s great. What ended up happening was, cancer was an amazing place to start all this work because of the absolute complexity of the manifestations of this disease. Two people– a doctor might say two people who have lung cancer, but they can be very different diseases, even though they’re in the same tissue. We have these names for them: non-small cell, small cell, whatever, etc. That heterogeneity inspired me to: “How do I optimally look at these patient populations to help make good clinical decisions?” Then, I did the same thing in neuropsychiatry and then I met one of my investors of NetraMark, the CEO of Juvenescence. He introduced me to some applications. 

Now, at the same time, when I was doing my cancer work with Igor Jurisica at the University of Toronto, there was an aging component. I was already playing with aging earlier on, I was trying to understand what happens during aging, how does this relate to cancer? There’s an intimate relationship between aging and cancer. It’s true as we get older there’s a higher chance, but it’s specifically because some of the machinery that starts to go wrong is pertinent for aging, specifically, the accumulation of errors. I can get into it. I started thinking about this even during my oncology training and then it was the CEO of Juvenescence that said, “Hey, this is what we’re doing. Your technology can be very powerful for this because you can literally relabel these complex patient populations, let the machine label it for us and then we can use machine learning and so forth. Then we develop the relationship on that front.” So, that was Greg Bailey, Dr. Greg Bailey, that pulled me back in.

Sufal Deb  9:13  

I want to ask, what exactly is NetraMark?

Joseph Geraci  9:16  

It’s a technology company that truly has an agnostic– it has a machine intelligence component technology in it that could be used for many things. We focused on pharmaceuticals because of my background in medicine and these complex datasets. Essentially what it is, it’s a pharma-tech company, essentially now, that allows us to form these minor miracles. This is what we do. We do the following things. A pharmaceutical company comes to us and says, “We are having trouble understanding this patient population, can we have access to your technology?” We give them access and what we do is we put their dataset inside of it. What the machine does is– we give them this thing called NetraAI, it’s very cool. 

It accelerates your research by up to 100 times. Literally, in 14 seconds, this thing can tell you about your– I’ve seen it do it and it got it exactly right. These patients are going to respond this way and these patients are going to respond this way. It’s very cool and it did it with very weak labelling. What it does is basically like a data microscope. It allows you to go in, zoom into the patients and say, “Hey, these people are together because of this reason and these people are together because of this reason.” In other words, what happens is you give it simple labels that say, for example, these people are responders, these people are non-responders. The machine will shatter even those subtypes into further subtypes. Just because someone’s a responder, they can be a responder for different reasons so that can influence your clinical trial in a major way. You can actually literally focus— it hands you a biomarker that says, “Hey, these people have a mechanism of response that matches your drug.” This can be used for drug response, placebo response, safety, profiling, all kinds of things. Basically, NetraMark is allowing us to stop relying on these blocky observational names we give to disease and trading them in for mechanistic profiles instead. We have set tools like that, so that’s one of our tools. 

The other tool literally is used to look. We’re having trouble using machine learning because there’s something going on with our data set, it’s overfitting and what we do is we allow– something we call Shatter. We allow it to pass through our system and it basically creates labels for your dataset. It relabels it. This is the future of AI. Everyone understands we can– if you have good data, you can use AI, it’s simple. We now have to understand the hard parts of the data, so that’s the part we’ve been starting to attack. How do you use intelligence, machine intelligence, to do this really difficult work of partitioning the data in such a way so that a gradient boost or a deep neural network can really do its job? Those are the two things. One is interactive augmented intelligence and the other one is like a big brother for machine learning. It’s like you want to train a kid to ride a bicycle, you stand next to it and you prepare it. That’s what some of our technology, that’s what Shatter is doing now.

Marvin Yan  12:43  

Oh, okay. Now, is my opportunity to ask about the Alzheimer’s paper I read, because, once again, this is not an AI podcast or machine learning podcast. I know you’ve been on one or two of those. One of the things that I remember you saying was that in the realm of medicine, there aren’t huge datasets. You want thousands, tens of thousands, but really, you’re getting, optimistically, a few 100 probably? 

Joseph Geraci  13:05  

That’s optimistic. Yes.

Marvin Yan  13:07  

Yeah, but I remember, in your Alzheimer’s paper, the dataset was actually pretty small as well, which is not surprising, of course. 

Joseph Geraci  13:13  

It was over 100 people. Yeah, yeah, yeah. 

Marvin Yan  13:15  

A few 100. I’m assuming that more data is always better, but I’m not sure if that’s true. I also don’t know if there’s a minimum amount of data you need. How is it that you’re able to get such results from such small datasets in the first place?

Joseph Geraci  13:29  

No, no, it’s a great question. Here it is. This is the way I’ve been thinking about statistics and machine learning. Statistics are great because you can gather confidence in an observation using statistics. Can I generate a significant p-value? Then say, “Okay, my observation is not just a fluke.” Okay, so then there are two aspects to machine learning from my perspective. 

One is the classic one, which is “I want to be able to build a model so I can make a prediction about an individual.” For those of you who can’t see, I’m holding up a pen. You show the pen to the AI, it says it’s a pen, it learned what a pen is. If I hold up a ball and it says it’s a ball. That’s a prediction. When you have smaller datasets, you have to be careful because it doesn’t gather enough of the distribution of what a pen is or what an Alzheimer’s patient is in order to make really accurate predictions. What you can do is you can extract. You can get the machine to create hypotheses about your data and what we did is we invented a mathematical paradigm that allows the machine to extract observations from the data that collectively is significant. In other words, it’s a hypothesis that you can really go out there and test or observe, or you can literally as a human say, “That’s crap or that’s junk.” 

I’ll give you an example. I’ll give you an example of crap and something to relate it to immortality, okay? Crap is that we got a perfect model for diabetes. Perfect! The machine handed us back two classes, it didn’t even make any other– it didn’t discover anything. It’s perfect right down to the number and we asked the machine, “Oh, cool what variables are you using?” Someone left in a variable called history of diabetes. The AI did its best and did perfectly because it was there. We were like, “Oh, crap, okay, that’s not good.” 

Here’s an example of– this is an amazing example, we put in aging data, that we have at NetraMark. It’s basically skin data sampled from a two-year-old up to a 96-year-old. It’s gene expression so it’s genetic and a very special type of genetic data. We asked the machine a stupid question, “Can you separate people who are over 50 from under 50?” What it did was that it basically created a data set that said, “Forget about all of these other people, they’re all mixed.” But, it pulled out the subpopulation. This is the magic, okay? It was able to ignore– it forgot. This is what me and some next-generation machine learning people are starting to talk about. Can you get a machine to be like a human? To forget, to ignore the noise. What it did is it yanked out a group that everybody– a pure group of over 50s or everybody was over 80, because it was able to recognize some signal that cohered all these people. No cheating, no internet, no labels, I just gave it a simple label, these people are over 50, these people are under 50. It said, “This is nonsense, except look at this hypothesis I just generated.” The p-value of those people coming together by fluke is like 10 to the negative 17. It’s going to allow you to generate a model based on the hypotheses that this thing can make because it knows how to forget and knows how to say “No, this is nonsense.” This is what’s necessary for the whole longevity space because the complexity is astounding when we age in different ways.

Sufal Deb  17:34  

Before I jump into the original question I was going to ask, I wanted to ask, is there a reason you use the word “forget” rather than “filtering out?” Is there a significant difference between the two?

Joseph Geraci  17:45  

I think it’s some of our desire that’s related to a lot of people in my space. You have to forgive me, I’m also a neuroscientist. We anthropomorphize. We love our tools. I think that– see, forgetting is a very important thing. We’re starting to refer to this process as forgetting, and you’re right, it is filtering. It’s basically saying that this is nonsense, but it’s a forgetting mechanism because what it does is– it’s forgetting that allows you to relearn, to recapture and not overfit. I use that term because I want to sound cool.

Sufal Deb  18:24  

Yeah, as somebody who works in a lab and has been working for quite some time now, I’m always told, any data is good data. Bad data is good data, more data is always good. Is more data always good with producing a predictive model? Is having more always necessarily beneficial?

Joseph Geraci  18:40  

Okay. Yeah. This is a beautiful question. The answer is simple. Yes, more data is better. But, there’s data that’s just garbage, it was not collected right, it’s contaminated. There’s data that comes from machines that change every hour. We literally had our machine intelligence create a hypothesis about a bunch of MRIs and the hypothesis was– it arranged everybody according to the time that they used their MRI in about 72-minute blocks because this thing was sliding for some reason. Something was wrong. Proteomics, same thing. You have to be careful. Yes, the more data the better simply because it’s a truth that phenomena in our universe are distributed. Mathematicians and statisticians get excited about distributions because they describe the way data is distributed. If you can figure out very precisely how distribution is formed or what it’s made out of, you can make really powerful predictions with that. The more data, the more full, the more complete the picture of the distribution. Unless it’s all garbage, then you’re not getting anything.

Marvin Yan  20:06  

Okay, before Sufal starts asking more questions about data and whether it’s good, bad, or how useful it is. We did touch on aging and I did want to elaborate more on it because when we talk about aging, depending on who you ask, some people are very strongly opposed to calling it a disease while other people are much on board with that. Joe, do you think we should define aging as a disease? Is it helpful in any sense?

Joseph Geraci  20:31  

Yes. It’s helpful because– maybe not a disease. It’s something like a disease. It’s a syndrome. It’s something we all approach because there’s an internal failing. Like, what is “dis-ease”? Right. Dr. Joe’s, as you know, if I put that hat on, it’s like, “Yeah, you’re messed up because something went wrong, and something’s broken, or there’s a genetic malfunction, or whatever.” What happens? Is it the– Okay, let’s look at cancer, okay? Cancer– let’s be logical. We accept that cancer is a disease and it causes severe dis-ease. Cancer occurs for various reasons, but at the bottom, it’s because there’s some sort of genetic malfunction that creates a neoplasm. Often what happens is the apoptotic signal breaks down, which is the suicide signal. That’s a very good example. All of the cigarette smokers out there, that’s what you’re doing to yourself. You’re disabling the ability for your cells to commit suicide when they have errors. As this error accumulates you end up with a neoplasm, a tumour. 

aging causes the same thing, you end up accumulating a bunch of errors, you end up with senescent cells, which release chemical signals you don’t want. You end up accumulating errors on your genes and so forth. Proteins don’t get translated properly. It’s like a radio station going out of tune. If I come along and I give you a drug that puts you back in tune, the music comes back clear then it seems like I’ve just cured some “dis-ease”. To me, it’s the same thing, just because this disease is inevitable, doesn’t mean it’s not a disease. That’s the problem, this inevitability. There’s spirituality attached to it. There’s all that. Which is fine, it’s great and it could be the most important aspect of our lives. We don’t know that. Death might be the most important. We don’t know what we are, we’re trapped between zero and 100. You can have spiritual practices and all that, that might infer something to you, but we don’t know what we are. All that aside, it’s an accumulation of errors that kills you. If not something else.

Sufal Deb  23:12  

Related to curing aging, or curing aging as a disease, there are two sides to the same coin, one being curing individual diseases that come from aging, or are associated with aging, versus just stopping people from aging altogether. Do you think one is better than the other? Do you think solving all these age-related diseases individually would lead to us stopping our aging process?

Joseph Geraci  23:35  

Yeah, so this question is important. We need to go after age-related diseases first because we can catalogue these things. Even Alzheimer’s is extremely heterogeneous, as you can see from my paper. We can attack those first. They’re in your face and they have to be dealt with. They’re the obvious things. aging, the mechanisms behind aging itself, is going to require, in my opinion, a next-generation therapeutic technology, like gene editing or something like that, or nanotechnology. For example, there are companies now experimenting with tiny little vacuums that literally vacuum up toxins from intracellular spaces, the spaces in between yourselves. This can change the way we age. I think that the progression is going to be natural. There are certain groups now, which are going directly after aging as a mechanism. I think that what we’re doing at NetraMark has to happen first, which is to create a map of disease. This goes back to your question of what we do. Really, that’s what we do. We’re creating maps of all these complex disorders, whether it’s lung cancer, aging or whatever. ALS. That’s what we’re doing. We’re creating these maps and we have to create a proper map of what aging– there are so many things that can go wrong. I believe in going after the individual diseases right now, yes. We might be able to even drug some of these things.

Sufal Deb  25:15  

If I can try to summarize what you said, do you think that we need to map out all these individual parts to be able to solve the bigger picture?

Joseph Geraci  25:22  

Yeah. 

Marvin Yan  25:24  

To follow up then, given what you learned about Alzheimer’s and that it’s heterogeneous, do you think we’re wrong about a lot of other age-related diseases, namely that they actually have multiple causes, rather than just one?

Joseph Geraci  25:35  

Yes, I do believe that. I believe it’s already damaged a lot of the work in this space. Thinking that one drug is going to cure Alzheimer’s, dementia, mild cognitive impairment, or age-related major depression. Inflammation is a big part of all this, but is it the only part? It’s a mess. It’s a combinatorial mess.

Sufal Deb  26:05  

I’ll switch gears completely and shift to a topic that’s been hot in the media and in research for quite some time, which is the idea of personalized medicine. Do you think health care, where it is today– obviously, every region is different, but in general the level of subgroups that we personalize healthcare to, is that enough? If not, how far should we be going with personalizing health care?

Joseph Geraci  26:26  

Yeah, it’s not being utilized properly. The motivation for doing things the way they are is because it makes money for certain groups of people. Right now, precision medicine, I believe, is going to be a critical aspect of the medicine of the future. Simply because one drug does not fit all. We’re not cutting things up to a granular enough perspective yet. I understand pharmaceutical companies want to maximize how much money they can make from a drug that they clear. But truly, this process is already happening. It’s already happening in cancer, where certain people have to have a gene expression over a certain level in order to get the drug and they throw away a significant part of their market to do this, but they pass the clinical trial. It’s starting to happen and no, we’re not utilizing that enough. Digital biomarkers, guys, this is going to become a huge part of it. There are going to be things that you’re going to be able to collect from your phone, from other things interacting with an app, with your watch that in combination is going to define your journey, your health journey throughout your life, if we can make this precise enough. This is where I’m starting to move towards. This is why I’m excited about joining Nurosene’s team because what we’re doing is we’re collecting data from the app and from other gadgets that the person’s wearing and we’re merging it with our understanding of disease. Eventually, it’s possible that a digital biomarker, a person that utilizes this in the proper way and puts in a lot of variables into this, might be able to identify subtypes of diseases that won’t happen for years. Thereby, really helping individuals. We can trade this stuff in for genetic information and all that eventually.

Marvin Yan  28:32  

In terms of the future of healthcare, I feel like I get what you’re saying, but there’s one point specifically, I know a lot of people will want me to explicitly ask, so I will ask it, which is about this fear of AI and all these models taking over the healthcare system. I think right now it’s more so where the doctor or the physician may use these tools to help make a decision, but ultimately, they’re using their medical expertise to make the final call. Do you see this ever changing where we no longer need physicians? Will what we call a physician now, will that role end up changing? What’s your thought on that?

Joseph Geraci  29:08  

I’ve learned to never say never. I’m not going to say, “Oh, no, never, we’re always going to want to rely on human doctors.” The power of a human physician is in their clinical experience, not in their creativity. It’s not in how well they understand physiology. They gather clinical experience. That’s why we take our kids to a physician, we trust them because they have experience. Now, machines are going to probably become smart enough one day to very accurately do this. The point here is cultural. How can we convince people to trust it? Is it possible that these children now– not the millennials or the generation before the millennials. Is it possible that their trust in technology, the way they interact constantly, is going to be the driving force behind exactly what you’re saying? If you ask a guy like me, I’m like, “No, never. No. Wow, we never know.” But if you ask my daughter, she’s gonna ask, “Yeah, well, how accurate is it?” She’s become like this. It’s become colder for her. She says, “Yeah, okay, maybe I’ll try it.” That’s why I do not say never. A company is going to probably develop some diagnostic system that’s going to be so good, but that’s gonna take time. Time, time, time, time, time. But it’s possible for certain things. The issue is data, how do we treat the data? I don’t know, I don’t know how to answer that question. No, I don’t know. You got me thinking too much, now.

Sufal Deb  30:57  

That’s all right. That’s the point of our podcast to get people thinking about these topics, somehow we got you thinking so, it’s an accomplishment in my book. Since you mentioned the word “spirituality”, I’ll make a very corny transition and ask, have you ever crossed paths with religion in your own life? In our podcast, we spoke to a lot of people who are very knowledgeable in certain religions and we’ve talked to them about their ideas about their religion or their religious experiences and what they think of immortality. For you, can you tell us a little bit if you have crossed paths with religion in your life?

Joseph Geraci  31:29  

I’ve always been attracted to Buddhism, Zen, Taoism and Hinduism. I’ve done quite a bit of study when I was a really young man on these topics. I even had the opportunity in my 30s, or young, early 30s, to sit with a Buddhist monk for a couple of years when I was living in Los Angeles. It was a transformative experience. I had all of these years of reading and practicing, in some sense, all of these different practices, including some Western mystical traditions. I was always attracted to it because my perception of science as a child didn’t split from my conception of nature, God, or reality. This is really me as a kid, I didn’t see any difference between God, a tree, and science. I was born a pantheist. I lived inside of God, so to speak. Then, it naturally evolved and then what happened is I spent years doing Buddhism, yoga, Jñāna yoga, Rāja yoga, studying Vivekananda and whoever else. When I started sitting with a Buddhist monk in Vipassana, I felt like all that scaffolding was torn away. All the scaffolding above God and above these– the naive stuff. I had an experience where for a few moments, I was– there was no ego-self. After passing through that experience, it altered me, my acceptance of death changed. That’s the answer. The answer is that I don’t want to die. I’m not insane. I still have– I’m an egomaniac but there’s an aspect of that experience that bled into how I perceive it, I’m a lot less stressed about it. 

What came out of that, if I may, is that I had this realization that there’s one consciousness, and that was split. This consciousness is split, like a fluid amongst vessels. A lot of our fear about death is just the vessel worrying about breaking, and not realizing that the fluids are going to pour into another vessel. I’ve spent a lot of time thinking about death. This is what happens when you’re a very curious kid who dives into Eastern mysticism. You end up realizing death is all religion. That’s why we have it because if we’re immortal, we’re gods. We’re terrified about what’s on the other side. I think it’s an extreme, maybe an equal part. People talk about Dr. Joe, all my degrees and all this stuff, but an equal part that’s hidden is this spiritual life where I really, really, really tried to focus on death. I would even read books that were meant for people who were dying. They knew they were going to die in six months. I put myself through those processes when I was quite young. So, I’m a mortal or immortal.

Sufal Deb  34:51  

With everything you’ve learned and everything we’ve discussed, do you think any of the goals of extending life, ending aging, or trying to become immortal, are out of line with your own visions and goals, and your own mantras?

Joseph Geraci  35:06  

No. No, no, no. It’s arrogant to think that you can become this immortal being. But it’s arrogant to think that the pursuit of this goal is not really, really worthy because, in the pursuit of it, we’re going to discover things that are going to help multitudes of people. That’s happening already. Starting to happen. These people that are focused on getting these mice to live seven times their age, we’re learning critical things about disease and it’s going to begin to drive therapeutics that go beyond our clumpy drug paradigm. That’s one of the great things about NetraMark. These maps I’m talking about, if you hand these maps to a genetic engineer of the future, they’re going to be like, “Oh, hey, there’s a subtype of ALS that we can just get rid of, by just manipulating these seven genes. Boom! Done! The nerve that’s retracting from the muscular junction, you can stop it!” 

Now, to me, it’s all the same thing. I just want to alleviate suffering. If I happen to live to 370, great! I’ll take on– I’ll become a dancer for a while, then I’ll become a scientist, then I’ll become whatever, you know what I’m saying? So no, it doesn’t– I don’t believe that just because I might have a belief that death is extremely important, it doesn’t mean we have to live but be trapped between 0 and 100. Maybe 250 would be a good number. Now, I meet people all the time that hate this idea. What they say to me is, “I want to be a healthy 85-year-old. I want to turn off at 90 and be done. Don’t want it anymore.” That’s the most common thing I hear. A few people are like, “No way. I want to live forever.” But, those people are usually very wealthy. Their perception changes about what life is like. That’s another— that’s a socioeconomic discussion.

Marvin Yan  37:17  

Okay, if other people share your sentiments, want to work on problems of aging and really want to make a difference. I know you’re somewhat of a polymath, like a man of the Renaissance involved in so many fields, but if someone asks you for advice, “Joe, I am really interested in this, I want to make a difference. Where should I start?” What would you say to them? 

Joseph Geraci  37:35  

Yeah, there are multiple paths. It depends on your predisposition. If you’re mathematical, I would recommend that you become a computer scientist with a bioinformatics background. Really get to understand machine learning and then dive deeply into biology. Then, find physicians to work with and create a company out of those relationships and that experience. If you’re not inclined mathematically and you can’t dive in at that level just because you don’t care, but you have it in you to become a medical doctor, then go there and then find the people I just described. This is how to do it. Not everyone has to be ridiculous, like me, and spend 30 years in university, but you need to build those bridges. I was on a call with a massive pharmaceutical company yesterday. He said “Yeah, we had this meeting but these data scientists didn’t want to talk to these other data scientists, and the physicians didn’t want to talk to these people. They all have their own silos. Joe, can you come and talk to everybody?” 

What happens is those silos have to break down. We need the physicians because we trust the physicians to gather data from our family members, from patients. We need computer scientists because they’re going to be able to do stuff with the data. We need mathematicians because they’re going to create novel ways of looking at the world. Mathematicians, physicists, whatever. This is the thing. Ultimately, I think it’s the diffusion of– let me be very general. Mathematics, computer science, and medical people– biologists. That’s the other way. Some people don’t have to be physicians, but they can become a biologist that specializes in how aging occurs, and they work with these people. I recommend that everybody learn how to program computers. Learn how to program R and Python. I’m not talking about you becoming like one of my engineers here. We know how to do DevOps and all this ridiculous software engineering stuff. Just being able to be competent from that perspective. Even physicians should have that. At least so they understand where things are coming from. That’s what I think we need to stop thinking we’re all separate. We’re all the same. It’s science.

Sufal Deb  39:58  

It’s like the idea of unity, you’ve got to put together ideas to come up with a result.

Joseph Geraci  40:02  

Yeah, right. It’s unification. I know people can’t see there’s a Rubik’s Cube. On the back of that Rubik’s Cube, those are Maxwell’s equations. This is a miracle. This is a unification of electricity and magnetism. This is where the universe opened up. It’s in unification, unifying things that seem like they’re apart where the magic occurred. As a theoretical– as a mathematical physicist, this is what really turns me on. This is where things become very beautiful.

Sufal Deb  40:37  

Joe, if there’s one thing you want people to take away from today’s conversation, what would it be?

Joseph Geraci  40:42  

It’s that they should invest in my company. No [laughs]. It’s that the path towards extending life is going to have a lot of effects for other diseases that we care about. It’s gonna affect our– there are implications in psychiatry, implications in oncology, and there are implications in neurodegeneration, massive. I think it’s a really great goal to get wrapped up in. I think longevity is an important goal because it’s like space or it’s like the Large Hadron Collider. This massively– “Okay, can we figure out? Is there the Higgs boson?” In order to figure that out we’ve had to invent– there had to be advances in metallurgy and in engineering that are going to have effects in other areas like medicine. This happens over and over. I think that this massive goal, aging x, is going to pull us towards a lot of discoveries about ourselves and improve our well-being, in many ways. Even though we might not hit that 1000 year lifespan that some people are talking about. Maybe we will.

Marvin Yan  42:03  

Well, for people who want to learn more about your work and maybe support it, possibly with dollars. How can they do so? Where can they go?

Joseph Geraci  42:12  

If they just Google Dr. Joseph Geraci NetraMark, you’ll find me. I’m all over the place. This piece has been written about me recently. I’m on LinkedIn. Look up, Joseph Geraci Nurosene, NetraMark. Nurosene is this great company that’s going to start– we’re just starting now to move to the next generation, which I was talking about earlier, which is using digital biomarkers to do all this great stuff we’re talking about would be great. It can become a very affordable way to do precision medicine. When we get there. That’s the way they can find me. Google me. 

Sufal Deb  42:57  

Perfect. For everyone listening, any links such as Nurosene or NetraMark will be down in the description below. Anything we discussed. Once again, thank you so much, Joseph, for coming on to Im a Mortal, your source for all things immortal. We really appreciate you taking the time to come and speak with us today.

Joseph Geraci  43:14  

Thank you. It was awesome.

[MUSIC – Im a Mortal Theme]