The Melanie Avalon Biohacking Podcast Episode #204 - Harry Glorikian

TRANSCRIPT 

Melanie Avalon: Hi, friends. Welcome back to the show. I am so incredibly excited about the conversation I'm about to have. It is about so many passions of mine. First of all, it's about something that we talk about on this show all the time, which is, health and wellness, and biohacking, surrounding that. But then it's about a second aspect, which is also something I am personally obsessed with, which is, the role of artificial intelligence, and what all of this actually looks like practically in the future. So, I am here today with Harry Glorikian. He is the host of a podcast called The Harry Glorikian podcast. He's also a global business expert, a healthcare entrepreneur, an author. His team reached out to me for his newest book called The Future You: How Artificial Intelligence Can Help You Get Healthier, Stress Less, and Live Longer. So, I saw that title and there were so many keywords, like I said, that I am just obsessed with. So, I was an immediate yes. 

I read the book and it's pretty mind blowing friends. We'll get into it in today's show, but it really just paints a picture of where we are headed when it comes to health and wellness and what that's going to look like. Well, now, I think to an extent that some people don't even realize and in the future, just as far as tracking our own health, taking charge of our own health, hospital systems, how that might manifest, patient care, things like genomic testing and gene editing. There's so much in here. So, I am just really, really excited about this conversation. So, Harry, thank you so much for being here. 

Harry Glorikian: Oh, thank you so much for having me. It's great to be here. 

Melanie Avalon: So, I have so many questions for you. But to start things off-- He's laughing, because we were talking before about how many questions' I sent him. To start things off though, you talk about your background in the book. Could you tell listeners a little bit about your background? What led you to doing what you're doing today? Have you always been interested in artificial intelligence and all this stuff? I will say, as a quick sidenote, you reference Star Trek a lot in the book and I'm obsessed with Star Trek. So, growing up, I was like a Trekkie from age 10 onward. So, I really appreciate that aspect of it. [giggles] But in any case, your background, what led you to doing everything that you're doing today with all of this work? 

Harry Glorikian: So, I've always been in the field of healthcare and life sciences. I started out my career in a startup that was doing pathology, or actually, we were doing, what I would consider, the first part of precision medicine. We were doing something called immunohistochemistry, where we could see the receptors on the cells, and able to help a pathologist figure out like, "Is a woman estrogen receptor positive or progesterone receptor positive?" And therefore, it would influence their treatment. Before that, you would basically grind up the tissue and use radioactive material, and they would come up with a centimole number and make a determination on that. For me, that was a wild guess, as opposed to a focused decision that you would make based on that next level of data. 

Then I worked in that space for a while. I had my own lab, I built it, I sold it. Then I went to go work for a company called Applied Biosystems. Very lucky, again, being at the bleeding edge of genomics. We were at the forefront finishing or doing the human genome through our sister company, Celera Genomics. We made all the instruments and everything to make that happen, which I have to tell you is still probably the finest time in my career if I had to pick a time. When we announced the genome, it's amazing to see thousands of people walking on air, because they're so super excited. 

I left there. I started a strategy consulting firm that I grew and I sold to a private equity fund. It was at the tail-- It was at the end of that where I was like, "Wait a minute, once they pass the Reinvestment and Recovery Act, and everybody is forced to adopt an electronic health record, everything is going to change from a data perspective." And so, I ended up writing my first book, MoneyBall Medicine, to capture areas that we're going to be fundamentally changed by data and data analytics. And then, of course, I've got the comment of, it's a little complex of what you've written, so can you write a future you for the average person, which I must admit was a little challenging, because you take a lot of information for granted. And so, I ended up writing that, publishing that. I also have my podcast where I interview people on the intersection of data and biology, and what they're doing and the impact that it's going to have. 

So, I've always been sorting at the bleeding edge of science and technology. I've been lucky enough. And then understanding about data and the impact of it, playing with it, and designing systems that that would hopefully answer that next question. And so, I thought, "Okay, this is definitely going to be a need for people to understand how this is going to impact them." And so, I play both sides of the fence. I play on where the science is truly happening. Since I'm in venture capital, I invest and interact with those people a lot. But at the same time, I think translating that to everybody else out there, so that they can understand how to take advantage of it is that role needs to be played too, as much as possible. 

Melanie Avalon: Well, first of all, I want to say it does not show that you struggled in making the book approachable, because it's very, very approachable. So, good job there. 

Harry Glorikian: I can thank my wife for that. 

Melanie Avalon: Awesome, awesome. Well, thank you to her, because yeah, it was very, very readable. You touched on so many things I would love to talk about. Well, speaking of that approachability aspect, so, you start off the book with a general discussion and explanation of what AI is and how it actually works and how it compares to a human mind. Can we just, as a foundation talk about that, especially right now, it's really timely too. When did you release this book? What year? 

Harry Glorikian: Not last December. The December before. 

Melanie Avalon: Okay. Were you worried? I know I asked a question. Now I'm asking another question. But [giggles] when you released a book like this, do you worry about being dated really fast because of how fast technology is progressing? 

Harry Glorikian: It depends on who the reader is. What I have found is that the vast majority of people are not current. Current is an understatement. Humans aren't designed to process at the rate that things are changing, which is part of the reason why I wanted to write this stuff, because I was like, "Uh-oh." [laughs] I have people who read the book and go, "So, this is going to happen in the future." I'm like, "No, no, everything in the book has already happened or is happening. It's just you have to look at it from, like you're looking at an evolutionary curve, except the curve is being crunched because of how fast everything is happening." So, in my last book, I have a quote, "The difference between evolution and revolution is time." 

Melanie Avalon: Oh, that's a good quote. [laughs] I like that. That completely makes sense. That's why even in the intro I was saying, it's the future, but it's also things happening now. Yeah, I really think a lot of people don't realize all of this is happening. Speaking of that timeline, how do you feel about this concept of the singularity and that AI will reach this point where it'll just exponentially grow and take over, really quick? 

Harry Glorikian: Skeptical. I'm watching, and very intrigued, and trying to keep up with what's happening on the GPT front. You can see how it's moving forward very fast. Now, if you really look under the hood, there's a lot of mathematical analysis and trickery going on. But you and I could have a discussion of what is truly human understanding, because we consume information, we process, and then "we have an understanding" based on the information that we've consumed. Sort of similar to what's happening on the GPT front. Does GPT understand purpose? Probably not. We understand purpose. So, there're some subtle differences between the way that we look at the world and do things and the way the machine does, right? But it's moving super-fast. I just don't see the AI sentient thing happening overnight. But if you think about something like ChatGPT, did they really know what they had when they released it in a chat format? I don't think so. I think they released it, and it blew up, and now they're running to try to keep up. 

Melanie Avalon: I'm so excited to be talking about this. Yeah, okay, because one of my major questions and I still want to provide a foundational explanation of how AI works, but a huge question I had, actually, was because you talk about in The Future You this concept of black box algorithms, and how AI and healthcare and things like that processes data will give an answer, and we don't know how it reached that answer. [giggles] And so, I was wondering if the parallel between that in healthcare and medicine versus something like ChatGPT, because my experience using ChatGPT, at first, it was really fun. I was asking it questions, fun things about [giggles] what it knew about me and my shows. Sometimes it was right, sometimes it was completely wrong. But what's interesting is the certainty with which it presents the information. 

So, I started using it to actually write some articles on a health supplement, one that I'm releasing. And so, you would ask it to write a scientific blog post on this supplement, it would provide references and a really great article, and I was like, "This is great." And so, then I went and actually read the articles it was referencing and it was just wrong. And so, then [laughs] I would ask it, can you provide from the study where you got that information? It would just be like, "Oh, I'm sorry, that was wrong." So, [laughs] it's very concerning, the level of certainty it presents this information with. All of that to say, how does that algorithm compare to in healthcare when AI is used to present answers with data about things? 

Harry Glorikian: Well, these large language models, there's a lot going on back there. I know I have not seen anyone be able to say, "Here are the exact steps and all the evidence--" Everything you need to do to be like, "Yep, I'm good. Yes, you've shown me enough. I believe you." Because if you think about it and you're like, "Write me an essay." In reality, if you just took the essay and used it, like, shame on you, right? It's, write me an essay or give me ideas and you should be able to take it from there. I find that ChatGPT is great at build me a four-day tour in San Francisco. I've got kids that are this age. It does a wonderful job. You still have to go in and fix a few things, but boy, it saved me an hour or so trying to design something for a friend of mine to go to San Francisco or summarize this for me and give me the highlights. It does a great job. Turn this into a Tweet or a LinkedIn post. It typically does a great job. 

I'm hearing from my friends in the coding world, GPT-4, it's really good. Is it perfect? No, but it's really good. Now I've heard GPT-5 is an order of magnitude better than GPT-4. So, that is a little scary, because if it's an order of magnitude of 4, you got to be thinking, "Do I really need half the people that I would call for help anymore?" I'm not so sure. 

Melanie Avalon: It's interesting, because I was thinking, one of my main concerns is just the level of certainty with which it presents this information. And then it doesn't tell you that it was wrong until you ask it. Literally, it's like talking to a psychopath was my experience, because it's very certain, very sure, and then you're like, "What about this?" And then it's like, "Oh, well, yeah, that was a lie." [laughs] It is like okay. So, I was thinking about the evolution through the versions. If it's getting more and more accurate, you're getting "safer in the way that information is presented." But what if there are still things that it's wrong about? But then I guess with humans, humans do the same thing. They think they're correct. I guess, maybe the responsibility, like you said, is more on us receiving the information and fact checking. But then the problem becomes, and I feel like this is me talking rather than asking you. [laughs] But then I wonder how the problem becomes when we no longer know what's being created by AI versus humans. 

Harry Glorikian: We're already there. 

Melanie Avalon: True. Do you think a lot of what we read on the internet now is that--?

Harry Glorikian: Not yet. I do believe that there's more and more that's being created as we speak. How far along it is? I don't know. You can imagine that in the next 12 months, there's going to be a lot of content out there that was created on one of these systems. There're already images that people are creating the art and posting it. 

Melanie Avalon: What's that word, that valley word? Do you know what I'm talking about? When it's artificial and it's mimicking human, but it's not quite there and so something is off. Do you know what I'm talking about? Something valley. Uncanny valley. Have you heard of uncanny valley? 

Harry Glorikian: No. 

Melanie Avalon: I would need to look it up more, but I was just wondering if that'll apply to words as well. I think it's the idea that, when you're so close to mimicking humans, but there's something slightly off, it creates this psychological experience called uncanny valley. So, that's a whole tangent. So, going back to the general overarching question of how this actually works. So, in the book, you do talk about artificial intelligence, and you talk about how there's different words that can be used to describe it. But how does it compare to the human brain? And humans, let's starting with the human. Is everything that we do artificial intelligence, or do we actually also do non--? How would I say this? Is everything that we do the way artificial intelligence is intended to function or do we also do very simple data processing that would not be the equivalent of artificial intelligence? Does that make sense? 

Harry Glorikian: Yeah. So, look, AI is inspired by the way the brain works, but it's not an exact mimic. If you think, like, there's chemistry, and electrical signals, and there're all sorts of things the way our brain does things that are not necessarily the same as an electronic system. The brain is complex. It's a biological organ. AI systems are designed to simulate some of the cognitive process and functions. When we were talking about some of these systems, AI mimics, the human brain is like a neural network, which is-- it is a mathematical model that simulates how we think of biological neurons. These networks are trained on large amounts of data, and they learn to recognize patterns, and then they make predictions based on that data. And so, that's why sometimes these models are big, because they're layers upon layers upon layers. One layer is making a decision and passing it to the next layer, and so on, and so forth. It's similar, again, to the way neurons in the human brain communicate with each other through synapses, but different. So, we're trying to use the brain as a model, but it doesn't exactly function the same way. 

Melanie Avalon: What's the mind behind it that's noticing all the patterns, or is it all just individual programs? What part of it is learning? I don't understand. [giggles] 

Harry Glorikian: So, these systems will use these mathematical models to identify patterns based on what it's seen. In other words, take the cat example, right? You keep showing it all the cats, and eventually, it figures out when it sees it in a YouTube video like, "Oh, that's cat. I recognize it. I see the ear, I see the tail, I see the nose. Okay, that's a cat." But if you really want it to be smart, you also want short dogs and other animals. And so, it can tell the difference between the house cat, and the dog, and the parrot, and everything else. It starts to create patterns where it understands. Understands, I hate that word because it's not an understanding. They recognize when something fits a particular pattern to say, this is X. So when we're, let's say, looking at MRI images, x-ray images, CT images, if the system has been trained to recognize something in that image, and that also has the diagnosis information with it, it can say, "Hey, I think this is X and the probability is it's this," because the image may be not perfect or whatever, and it may look like two different things, and it says, "Okay, this one is probably the highest probability, and this one is the one below it. Mr. Doctor, you've got the experience, you should be able to determine which one it is." But with a pretty high certainty of probability, it's one of these two or three. You see what I'm saying? 

Melanie Avalon: Mm-hmm.

Harry Glorikian: It can recognize it. But we're pretty good at this. Remember, we've been looking at satellite images and looking for nuclear missile silos for a long time.

Melanie Avalon: With that recognition, you talk in the book about how health conditions, where we take a note of data, is very ripe for AI, so things like cancer and heart disease, is it just having the large amount of data that really helps or some diseases more patternable than others? Are there some elusive diseases that just don't seem to have a pattern and AI struggles to figure them out even, or is it all going to be a pattern in the end, you think?

Harry Glorikian: I think if you gave it the right data, the probability of it triggering out is high. I say that only because you see it do things every once in a while or highlight something and you're like, "Huh, would have never come up with that on my own. Like, it just wouldn't have happened." It noticed a pattern that I just wouldn't have noticed. Most people wouldn't have noticed. The other thing is it's chewing on so much data, I think the human brain is not designed to see-- can't see that pattern at that level of detail. We always say, a picture is worth a thousand words. That's our way of sometimes seeing a pattern. We put the picture together, we're like, "Uh-huh, the line is moving in this direction." 

Now imagine the machine doing that with 20 lights, 30 lights, 50 lights, 100 lights. And so, it can see-- Oh, wait a minute, but in these 100, these 5 always do this, these 10 always do that. And so, it can start to subsegment, and see things that you wouldn't, because we're just not designed to do that. There's Etiometry, which is a local Boston company here. They have a system that sits in the intensive care unit. Now, if you've ever been in an intensive care unit, there's a lot of instruments in there, and a lot of beeps and boops, it's so much where a human being just isn't going to notice all the beeps and the boops and how they're all interacting. But their system will be able to alert someone and say, "Hey, listen, in the next, I don't know, 24 hours, 48 hours, the patient in bed number 3 is going to have a problem. You should go in and do something." And so, you go in there and you intervene. But if you didn't have that early warning system, am I going to notice the patient number 3 bed, all the beeps in the boops, do the analysis, see the outcome? Yup, I need to do something now, right? 

We're just not designed to that. So, it can be an amazing tool to help us understand where we need to focus our attention sometimes. And there're dozens of these examples. There's a, I forget who it was. I want to say it was maybe Butterfly Network with a portable ultrasound, it'll say like, "No, you can give it to the average person. Move a little to the left, move a little up, move up, no, move down, move--" And it tells you how to manipulate the ultrasound sensor for you to get an image that's equivalent to a trained technician, because the system in the background is constantly looking, doing, it gets calculations, recalculating saying, "No, that's not good enough. I need you to fill in this gap for me by going up to the left." In the end, you end up with a good image. So, there're a lot of ways that the technology will help us. I need to make this molecule. Can the system suggest a different way of making it, maybe faster, maybe cheaper, maybe I'm looking for a different property, whatever it is. But the system will look at all options. It may suggest something to the person that was like, "Wait, I never learned that in school. They didn't teach me that. I would have never come up with that pathway." But the machine doesn't have that limitation. 

Melanie Avalon: I would like to see a poll of everybody, how they feel about doctor only care with all of this. And then doctor plus AI and then just AI. You do talk a lot in the book about, will AI render doctors obsolete? It's so true. Going back to Star Trek, my love. I'm assuming you like Star Trek? 

Harry Glorikian: Love it. 

Melanie Avalon: Which series? 

Harry Glorikian: Oh, the original one. More or less anything with Picard got me too. But a lot of the other ones, I tried to get into Deep Space Nine and all the way. But anything with Picard in it or the original and I can still watch the original today. 

Melanie Avalon: Oh, me too. Oh, and I do. I do.

Harry Glorikian: So, here's the difference. If you went back early enough, what did a doctor do? There was a limited amount of things a doctor could do for you, because the technology wasn't there. And so, you got a lot of personal attention. Most doctors knew you from cradle to grave to a certain degree. Then technology comes along and, oh, my God, we can do everything. Well, now that person has to understand all these different things, how to apply them, how to use them, what do they mean? Okay, you're starting to ask for a lot from a human and sorry, all doctors aren't on the upper echelon of the bell curve, right? It's still a bell curve. And so, you're expecting more from them? Well, but now there's an introduction of a technology that can help, focus, or identify things for them that it's just an assistant that's helping them with what they're going to do next. 

So, if you look at a lot of the papers that are coming out, what I typically see is the machine scores X, the doctor scores Y. Maybe one is a little bit better than the other depending on the area. But when you put the two together, they outscore either one on their own. So, man plus machine ends up scoring higher. Do I want my physician poring over data? No, I want my physician spending time with me. So, if the system helps identify the problem and/or the solution and we can get there faster, I'm all over it. A perfect example is Geisinger has a system that will look at brain scans. If it sees a bleed in the brain, it will take that scan and actually move it to the front of the line, so that the physician sees that first, so that they can do something about it [chuckles] rather than let that person have that problem until they get there in sequential order. And so, if you're that person, you'd be like, "Oh, my God, thank God, you guys had that system in place," because we have a saying that, "Time is tissue." If you don't treat the problem at that moment, you see more tissue damage. 

Melanie Avalon: The reason I was framing it with the Star Trek reference was it never occurred to me with that show. I am guessing that was a conscious decision to make the Dr. McCoy very human. He's just an old country doctor. You pointed out that, because Star Trek was so ahead of its time, and it's very spot on with a lot of its predictions, and even there in that show there's still a doctor, they didn't have the medical aspect completely replaced by computer technology, like, there is still a human involved, which I think is very telling. 

Harry Glorikian: Yeah. Will we be able to automate certain things? Probably. Do you still want the human to validate it? Absolutely. [laughs] Nobody's going to be happy with the computer giving them the bad news or coaching them on a treatment path right out of the gate. It may help them downstream to stay on the path, but not in the beginning. If you're in Sub-Saharan Africa, and there's a portable ultrasound, and the ultrasound helps that person take a better image, that's pretty helpful, where the doctor can't be there. Or there're some systems I'm looking at that are in the operating room where you can put the tumor into the machine, it would then take a slice and take an image, and then the image would automatically go to the pathologist who needs to look at it.

Normally, you'd have the pathologist in the operating room, but we have a shortage of those pathologists. So, if the machine can take that place and fill that gap, and then show the pathologist the image and then highlight the parts in the image that the pathologist may want to focus on, everybody wins. 

Melanie Avalon: Yeah. It seems like giving you a flashlight. I guess, it's like a flashlight that also might be wrong sometimes. So, you have to [laughs] fact check. 

Harry Glorikian: Look, so, here's the difference between GPT and medicine. In GPT, if it's wrong, okay, hopefully the world is not coming to an end. If you were making that important of a decision, you should have been a little bit more diligent on what you were doing. In the world of healthcare, we have clinical trials, FDA approval, people scrutinizing it. We typically have to jump through a lot of hoops to get something out there. And then, you got to get everybody to adopt it. While they're adopting it, they're playing with it too. Whereas GPT, everybody's trying to spoof it or get it to do something wrong or things you can't just ask it any question you want. Is it designed to do a specific function over and over and over again consistently? 

So, the bar is higher. I'm not saying that there couldn't be a problem, let's say, in its early days, but we're trying to do everything we can to eliminate that, minimize it, because people's lives are at stake. And then the fail-safe is the physician, to take a look at this and say, yes, no, and then move forward from there. 

Melanie Avalon: What is the role, because you were just speaking about how basically a person getting bumped to the front of the line having more of an issue. And you talk all throughout the book about how this can change the hospital experience, and ER waiting rooms, and things like that. What is the role of bias in these calculations? I'm really fascinated by how things like racial bias, for example, that they find that happening sometimes in all the different programs and modalities? Is that just because of the data set that's used? Why does that happen? 

Harry Glorikian: Yeah, some of it is most of the time, it's the data set that the system is being trained on. These models are only as good as the data they're trained on. So, if you give it a biased data set, guess what? I'm sorry. If you train your kid a certain way, they end up a certain way too. So, it's not dissimilar in that sense of you want to give it a representative data set that it can be trained on, so it's not underrepresenting or overrepresenting a certain group of people that therefore it can give you a better answer. So, every time there's a problem, there's somebody that's willing to sell you something that's going to solve your problem. So, I've seen a lot of these systems now where they're using encryption technology to bring in broader data sets from more areas without making the data accessible, therefore being able to train the model on a broader data set. 

Even if a hospital took a model from another hospital from a different area, they would still want to test it out on their population. In my population is it giving me the answers I think I should be getting? But as I said, if you're going through FDA approval and you're going through a clinical trial and so forth, you will be held to a higher standard and they're going to ask you in your trial, "Who made up your population and is it representative?" before they just let you get cleared and be released into the wild.

Melanie Avalon: I feel like I'm asking about all the fears concerning this, but you just mentioned using encrypted data from other sources, what is the concern if the data is completely anonymous? I feel like this is a naive question, but I've wondered this, like, what does it matter if my data--? Even if it does get in the hands of somebody who shouldn't have it, how does that affect me if it's anonymous? 

Harry Glorikian: Well, if I do figure out it's you, that's a problem. Well, first of all, the laws in this country, the United States specifically, just our lawmakers are so far behind. It's not even funny how far behind they are. They can't tell the difference between Facebook and Google let alone any of the stuff that I'm talking about. So, that is your first and foremost problem. Because you got to have the laws and the systems in place, so that everybody gets to a different level of taking some of this stuff seriously. Look, I can go to Singapore, stick my ATM card in and take out money, and it's secure. I can't go from one hospital to the other and get my data. That's just not possible. It was never designed that way from the ground up, which is mind boggling to me that it's still like that to this day, and it won't change until it gets legislated that they need to change it. 

Let me give you an example. So, I was talking to a company the other day that actually has come up with a cancer test for dogs. We were talking and there is no HIPAA for dog date. So, she's like, "Harry, we've got it all." I'm like, "Oh." "Yep, all everything." I'm like, "Ah, wow, you could do some really interesting--" And she's like, "Yep." She's like, "We can do some analysis, and we can see things, and we can move forward faster than anybody can in the human world," because I have to get permission for every little thing. You can see what they're finding, the biomarkers, which biomarkers are the exact same ones that are in humans are showing up in dogs. So, that drug actually will work on that dog. Or, you see what I'm saying. All of a sudden it just changes what you can do when the data becomes available. You almost wish that we could aggregate the data in the United States to move the ball forward faster. There're philosophical reasons that you may or may not want to do that, but from a pure science perspective, the larger the N of data points you have, the faster your research can move and where you'll see things that you may not have seen before. 

Melanie Avalon: I actually have a huge question about that, because you talk in the book about virtual clinical trials or using AI to, yeah, basically to work with clinical trials, and this concept of digital twins, so you can get more data without actually needing more people. And then you talk about how it can beneficial, because-- Well, on top of that, so the digital twins concept-- Oh, wait, that's actually two different questions. Okay. [giggles] So, starting with the digital twins concept, does creating all of that extra data, does that account for an actual data set that might look different from that? It seems like it would miss some things. 

Harry Glorikian: Well, I mean, it shouldn't. So, if you go into industry, this whole concept of digital twins is very normal. I create a digital replica of my engine and now I can test it all sorts of different ways to figure out what its tolerance is. If I do something to it, will it react a different way? But I can play with it in the digital world without having to do it in the real world. I can get all sorts of data from there that helps me make the one in the real world that much better. So, can I do that with humans or human patients? And the answer is yes. And so, I can create a digital replica of this person. Oh, by the way, the FDA accepts digital twins as part of your clinical trial. Now, you can't do 100% of the people that way, at least not yet. But let's say, you need 100 people in your trial, and you just barely got to 80. [chuckles] You worked really hard, you got your 80, but you still need another 20. Could you supplement 20 people in your trial? And the answer is, probably yes. And so, you can now start your trial and run your trial, but 20 of them are digital twins of the model patient. And so, when you do what you do, that digital twin should mimic the population of real patient. 

Melanie Avalon: I guess, it would depend on the setup of the study. But would that digital twin presumably be used as a control person in the study rather than the experimental or can it be either side? 

Harry Glorikian: It could be either side. It should be, you shouldn't say, one is going to fall in and one is going to fall in the other, because then you're treating one population different than you are everything else. 

Melanie Avalon: I find this so interesting. And then what about the second related question? Because you talk about how AI can help and it relates to what you just said about doing a clinical trial and not getting enough people to do the trial, how AI can help with patient selection, and finding people that may be more appropriate for the trial and may respond better and less side effects. But wouldn't that bias the trial to support, if it's creating a drug? Wouldn't that bias the trial to support the drug it's creating if you're using AI to select patients that will react favorably in the trial? 

Harry Glorikian: No. So, if you look at something like Alzheimer's, the number of failures is significant. Most of it, I believe, is because we can't subselect for the population, because we don't have a good way of substratifying the population into the right beings. But if you know that you can get to the right population that fits your trial, that would benefit from this drug, or you want to prove that your drug will benefit this population, you'd like to start with the right people to begin with. You don't want to give the drug to the absolute wrong people, do you? 

Melanie Avalon: I guess, it depends on the endpoint. When the drug is created, is it going to be mandated for that population that was the population that created it, rather than just Alzheimer's overall?

Harry Glorikian: No, it wouldn't, because these days most of the drugs that are being released are, you got to fit in this little box and then we prescribe it to you. 

Melanie Avalon: That's more approachable. 

Harry Glorikian: Yeah, yeah, yeah, no, no, no. And the thing is, the days of, like, I'm going to give you a drug and the drugs are going to work on everybody, those are few and far between these days. Whereas what you're finding is-- If you go back in time, when somebody had breast cancers, they had breast cancer. That was it. It was a big umbrella. Everybody fit under the umbrella. But no, as time has gone on no, there're all these substratifications of breast cancer. What specific form of breast cancer do you have? And then the treatment paradigm for that would be different than one of the other ones. And so, just like we say, everybody's different. When you call one of these diseases, typically, it's not one form. There're multiple forms of that disease that we either know it and we continue to dig down deeper or we don't know it and we're finding out as we go along. 

Melanie Avalon: I think that was the key piece of information I needed to understand why that's not concerning to me. I have no idea about this. Was AI involved in speaking of Alzheimer's? I should know more about this, but the whole scandal about the drugs not being based on falsified data, was AI involved in that discovery? 

Harry Glorikian: I have no idea. 

Melanie Avalon: Do you know what I'm talking about though, the recent-- [crosstalk] 

Harry Glorikian: Yes. 

Melanie Avalon: Okay. I'm going to google. I'm going to look into that. It just occurred to me. I think my closest experience-- Well, you're talking about how it's everywhere and we don't even realize it, so I've probably been involved in it a lot. But my own personal in my home experience has been, I have an Oura Ring. When I got COVID, it was so interesting to A, it did predict it sort of. It didn't say COVID, but it knew I was not feeling well. And then I could really track my progression on the ring. It was really cool to see. Do you use any wearables yourself? 

Harry Glorikian: Oh, yes, actually, I had the Oura. I have my Apple Watch. I have a WHOOP. I had tried the Levels CGM. I had tried the January's CGM. What else am I missing? I have a Withings scale, I have the Withings blood pressure cuff. I try and play with as many of these as I can. I have the AliveCor ECG. So, I try to play with as many of these as I can to see the Eight Sleep bed now that I'm thinking about it. I'm sure if my wife and my kids were here, they would remind me about another something that I was probably playing with at some point. 

Melanie Avalon: I actually have the Eight Sleep. I don't use it because it requires turning on the Wi-Fi at night and I like to turn off my Wi-Fi. So, I keep waiting. I need to reach out to them again. They were saying that it might update to not need the Wi-Fi in the future. I've had it since the beginning, since it came out. Just been chilling on the bed. No pun intended. [giggles] So, I hope to turn it on someday. In the meantime, I've just been using OOLER. So, going back to something you mentioned in the beginning, which was, your involvement in the human genome and that excitement. One question to start off and you talk about this in the book, but why did we think there were way more genes than there ended up being? Didn't we originally think it was millions and then it was less and less and less? 

Harry Glorikian: Well, you don't know what you don't know before you go and poke around in an area. There're so many examples of that where I've been with incredibly brilliant people. They make some declarative statement and you're scratching your head going, "All right, they're the experts. Theoretically, they should know." And then once you dig into it, you'd be like, "Nope, nope, [chuckles] that's not how it happened." 

Melanie Avalon: Not accurate. 

Harry Glorikian: I have stories of when I was at AVI, and people would say something, and then you'd find out later about. We dug into that and that's not exactly how it came out. 

Melanie Avalon: Everything happening with that-- It's really cool. My brother's fiancée is actually-- She just graduated. But her degree is in genetic counseling. So, she's currently looking for a job in that sphere. So, what do you see with that at present and the future of that as far as doing these genomic tests for newborns, and looking at potential risks, and addressing that--? Practically, where do you think that's going and what are your thoughts on the ethics surrounding it? 

Harry Glorikian: So, there's been a lot of work done on this out of Harvard, where they've actually asked parents, "Do you want to know? Are you glad you knew everything?" It always comes out with, "I'm glad I knew, I'm glad I found out. I haven't seen the other side of it necessarily. Where is this technology?" The impact is huge, huge in a sense of, I don't think most people appreciate how impactful it can be. If you have a child that is born and something's wrong, as a parent you're going nuts trying to figure out what's wrong with your kid. It's funny because if it was yourself, you might not be that involved, but when it's your kid, you're trying to move mountains, trying to figure out what's going on. 

In the past, we have a word called the diagnostic odyssey. It could have taken years to figure out what's going on, whereas what we're seeing is when you can do a whole genome sequence. I think the world's record now is like-- I should actually know, because I interviewed the CEO of Rady Children's Hospital in San Diego and they hold the record. I think it's seven hours from sample to result to identify the issue with the child, and be able to do something. That's huge. I want that. And I don't want it to take seven hours. I actually want it to be done in two hours, which we're going to get there, especially with AI on the backend, helping accelerate the analysis and also point out what the treatment should probably be. But the UK is now sequencing, I believe it's a million newborns. If not all newborns I can't remember. And if you can identify a problem you can fix, you can fix it up front and that child will never have to deal with it again. 

Melanie Avalon: What are some practical examples of things that's finding, just for listeners? Because you talk about the different-- I think there're three tiers of results for newborns. How expansive is this? What is this actually testing for? 

Harry Glorikian: So, when you're looking at the child, there's multiple levels of newborn screening tests. There's everything from hearing loss, congenital heart defects, metabolic disorders. And so, you see things as far as from hearing screening, blood spot screening, which all newborns get although if you go state to state, there are some differences. But when you're doing these genomics tests, you can see where you might be able to see an issue with the heart. Well, that's a mechanical thing. Can I go in and fix that mechanical problem? Before, actually, a problem happens later on in life, can I go and fix the problem now? And the answer is, most of the time, I don't think I've heard a parent say, "No, they'll fix that." There might be religious reasons why they say no, which I might disagree with, but you'd want to go in and make sure your baby is as healthy as possible. If you can fix something, you'll go in there and fix it. 

One of the examples that I have in the book though is a little different. You sequence the baby, and they come and they call up the mom, and they're like, "We have good news and bad news. We found something. Your baby boy is BRCA positive." "Well, okay, but that's not going to technically should not affect the baby boy, although men do develop breast cancer." But he's a carrier and they'll pass it on, but that means that they got it from someplace. So, we should sequence all the females in the family. No, by the way, [unintelligible [00:50:14] BRCA positive. And so, I remember asking her, "Are you happy?" She goes, "Oh, totally. I rather know about this and manage this upfront than have it sneak up on me and it be too late." So, she would not have known had they not sequenced the baby.

Melanie Avalon: That's another poll I would like to see. I personally would want to know.

Harry Glorikian: If you want to see these-- You're calling it a poll. I've seen these in full blown formal studies out of Robert Green's lab at Harvard. If you google him, you can see the papers that have been published literally asking and answering the questions that you have. 

Melanie Avalon: Awesome. Thank you. I will do that. Does he poll about, if there was predictive analysis for mortality, like personal mortality? 

Harry Glorikian: I don't know all the questions that he's asked, but I know that everybody has pulled and pushed on him in a lot of different, like, "What you're doing is unethical. People don't want to know. Parents aren't ready." He's gone out there and done formal studies where they really asked a lot of people. I think he's been on the right side of this all the time. 

Melanie Avalon: I will definitely look up his work. And for listeners, I'll put links to it in the show notes. Personal question for you. I'm always really curious by this. Would you want to live forever? 

Harry Glorikian: Did you see the Good Life? 

Melanie Avalon: No. 

Harry Glorikian: So, I think it was Netflix, the Good Life. You basically lived forever in what was supposed to be heaven, but let's not go into the premise of this show. It's actually quite funny, so I would recommend it. But no, I don't think so. I'd like to learn for the rest of my life. I got to be honest with you, I keep threatening to go back to school, but I don't think I'd want to live forever. That would be [unintelligible [00:52:12], I think I'd get bored after a while. And then you'd keep watching humans make the same mistakes, you'd be like, "Guys, can't you learn?" But I don't think we were designed for that. If I was super healthy, when I want to live for a good period of time, yeah, I'm having trouble seeing 100. I'm like, "Oh, my God." Because aging is tough when you're not in good health. I guess, the short answer is no, but I don't know what the age I would put on saying, "Okay, I'm done." It's not 75, I can tell you that, but I don't think it's 120 either. 

Melanie Avalon: Actually, you're definitely in the majority, I think, because I ask a lot of guests on this show that question, because I always thought, I don't know, I think I want to live forever. I think I always thought most people did, but it's hard for me to find somebody who says they want to-- Do you know Sergey Young? He's kind of in your sphere. 

Harry Glorikian: That sounds familiar. 

Melanie Avalon: He does similar to you. He invests in, well, longevity related technology, but his book is called The Science and Technology of Growing Young. So, he overlaps with a lot of the names that you talk about throughout your book. But he talks about in his book at the beginning-- Was he at the Vatican or something? He was somewhere on a seminar on aging, I think. I'm probably telling the story wrong. Regardless, it was a room of a lot of people in this sphere, and they asked, who would want to live forever? And he talked about how so few people raised their hand. But he's all about it. [laughs] That's his mission. So, I'm always just really, really interested in people's thoughts on that. 

Harry Glorikian: Well, he's a longevity investor, so I hope he believes that. Otherwise, getting people to invest in your fund might be a little challenging if you don't want to live by what you're doing. 

Melanie Avalon: It might be helpful. Because you talk a lot about 3D printing, and implantables, and all of this stuff. Do you think we will one day be able to have a transfer consciousness to a bionic body?

Harry Glorikian: I hope not. I don't think so, I don't know, I believe that there's chemistry that goes on and that chemistry has a lot to do with who you are. You're made up of all your experiences. I don't know how you would just port that over, if you know what I mean. I believe that when we go through all of our experiences, the codes that are written, it's not a hard code necessarily like a computer code. I believe it's a malleable code that changes over time. And so, some memories are stronger, some memories are weaker. There's a lot of chemistry there. So, I don't know how you would transfer that over and then have it be the same and then have it evolve over time the way that a human would evolve over time. It might be, who knows what's going to happen 100 years from now, but I don't see a logical path to get there per se. 

Melanie Avalon: The quote logical path I would see would be, I don't know if we could ever do this, but replacing one thing at a time in a person until-- So, rather than transferring it over, replacing piece by piece, cell by cell, so then they're eventually just all artificial. 

Harry Glorikian: Right. But if you think about it, it's the differences in things that make you you, your microbiome, your liver, your brain. If somebody tells me like, "We're going to find normal," I'm like, "You guys are-- That's just walk. There's no normal. There's a spectrum." We are meant to evolve. We don't stay static. If we were static, we would not be where we are now. And so, if you create something fully artificial, how does that evolve over time? If it's static or normalized, then I don't know how you become you. Now, if you need an artificial pancreas or something like that, I get it. It's doing a process and you want it to do that process well. But what makes up the whole system, I think, needs to be given room to flex. 

Melanie Avalon: Yeah. I guess, the way it would have to evolve would be the central-- control center, mind aspect, intelligence evolving. And then the intelligence would have to go in and actually upgrade itself. So, it has to be an external evolution imposed upon itself from the internal evolution of the mind. Like, the actual internal mechanics couldn't evolve themselves naturally, if that makes sense. 

Harry Glorikian: Look, sometimes I'm like, "Oh, my God, I really just want some pretzels." And you're like, "Okay, let's think about this, Harry." Is that me wanting the pretzel, or is it the microbiome pushing you to want a pretzel or the sweet or the this or the that? We're such an integrated system. When I say microbiome, we're not on our own, because those are different bugs that we're living symbiotically with. To recreate that whole thing, if somebody says, "That's going to happen," I would be like, "You're being naive. You're being incredibly naive on the complexity of what you're trying to do." 

Melanie Avalon: Have you worked with AI-related companies with the microbiome? 

Harry Glorikian: Yeah, I've tried to work with companies in the AI space in almost every space that I could touch or talk to. 

Melanie Avalon: What's your favorite space? 

Harry Glorikian: I will be biased and say things like, the genome and so forth, I'm biased because I do believe it has such a fundamental impact on everything. Anything that has DNA, it will have some level of impact on. I don't know how a country goes forward without that fundamental capability on multiple levels. So, that would be one. I would say that the sensor technologies are also quite exciting on how you can monitor different things all at the same time, and you can see the system in the background being able to say something's wrong. You need to have this checked out, or accurately be able to diagnose you in some way, or encourage you to do something that would be positive to you. If you tell me where the future of healthcare medicine is going, I believe I would say that these technologies can be early warning systems and coaches to keep you healthy or encourage you to get healthier.

Melanie Avalon: If a person has done something 23andMe or me for example, years ago, this was years ago, do you remember Genes for Good on Facebook? 

Harry Glorikian: I don't remember that one. 

Melanie Avalon: It was on Facebook and it was a project where they would basically give you the equivalent of 23andMe, but it was free because they were looking for research. They were doing research. So, I did that. Again, naive question, if you've sequenced your genome once in the past, do you need to do it again? Has anything changed? If I did it now, would it come out any different or could it come out different? 

Harry Glorikian: If you're exposed to different chemicals in your environment and things like that, you can have mutations occur over time. 

Melanie Avalon: So, the technology itself for sequencing, does it evolve much or is it set and now it's more--?

Harry Glorikian: Oh my God, it's been evolving like crazy. We have beat the pants off of Moore's law. 

Melanie Avalon: Wow. What do you recommend for the best resource for people who want to sequence their genome at home? You don't have to make a recommendation, if you don't feel comfortable. 

Harry Glorikian: A real medical level genome? 

Melanie Avalon: Yeah. 

Harry Glorikian: Okay. So, there's this line I have like 23andMe, I don't categorize as a true medical genome. 

Melanie Avalon: Okay, that's good to know. 

Harry Glorikian: If you're looking for a long lost brother or something or there are certain things that they look at, yes. 

Melanie Avalon: Is the data there, they just don't give it to you or is the data not even there? 

Harry Glorikian: You mean, when you get your sequence information? 

Melanie Avalon: Yes. I know when I did Genes for Good, it gave me the raw data and then it gave me some information, but then I could run it through Prometheus and go down the rabbit hole. 

Harry Glorikian: So, I don't know what level of detail that Genes for Good does. But when I did my genome, I went straight to Illumina, which is one of the sequencing companies. They had a program where they would do your gene sequence, give you information on what was a validated marker. If it was not validated, they didn't tell you. They would give your data, and you could opt out of sharing it with anybody else, so you didn't have to give it up to someone. And so, that's the one I did. I fortunately have an incredibly boring genome, which in most circumstances, you never want to be boring. But this one, this is a good one, you want to be boring. I learned a couple of things. If I'm on a long flight, I should get up and walk around more, because I have a higher propensity to clot in my lower extremities or I take an aspirin before I get on a long flight, as well as walk around. So, I've learned some precautionary things. 

I've understood what drugs I metabolize and don't metabolize well. There's a bunch of them, I don't think I will ever, ever, ever take those drugs, but they're on the list nonetheless. But this is a medical grade level genome. So, if people want to go get a medical grade genome, if there's a hospital in there, wherever they are physically that does a genomic analysis, they could go there. I know Harvard has a lab. I believe Bob Green runs that lab where they do whole genome analysis for individuals. A lot of the other ones that are available to consumers, I just don't categorize those as true medical quality level genomic analysis. 

Melanie Avalon: Yeah, I know the Genes for Good data set. It gave me two. It gave me the normal one and it gave me an imputed one. I was always really on the fence about, do I look at the one where it's filling in the data or do I not? I got really excited seeing a psychologist once and he told me that I could do a genomic test to see how I react to drugs. I was so excited. I was like, "Yes, please, sign me up." So, yes, all of this is just so fascinating. Well, I want to be really respectful of your time. One last topic that you touch on in the book, and I realize if we talk about this, I'm going to get the episode is going to get flagged. But [laughs] you do have-- [crosstalk] 

Harry Glorikian: Really? Okay. [laughs] 

Melanie Avalon: Oh, yeah. Well, you have a whole section on vaccines and COVID, which was very enlightening, especially the education you provided surrounding how the vaccine was seemingly "rushed to market for COVID." I was just wondering if you could touch briefly on the evolution of AI with vaccines, and what went down with that? If people's concerns are valid or not, what the future might be with that?

Harry Glorikian: Here's one of the big differences. When we've developed things in the past, everything happens serially for the most part. I do step A, then I do step B, then I do step C. Okay. But you know as well as I do, you do things serially, they just take a long time. During COVID, people took unbelievable levels of risk, and they did a bunch of stuff in parallel. Under normal circumstances, if you were developing a drug, you would not be building the manufacturing plant for the drug until you know the drug works, because as CEO would you be like, "Oh, I'm going to go spend, I don't know, hundreds of millions of dollars, and just I don't know if it's going to work." No, absolutely not. You'd be fired as CEO because you took unbelievable risk. 

Whereas during COVID, the government said, "We will give you the money. Please build the manufacturing plant. We're hoping it works. If it works, the whole thing comes together at one time." And so, there were a lot of things happening in parallel, because we were at war with this virus. So, the normal course of do step one, then do step two, then do step three, they were all happening at the same time. So, you took this, what would normally have taken five years at a minimum, and you crunched it down to about nine months. 

Now, on top of that, you are sequencing COVID. Well, I can tell you that when we sequenced SARS at Applied Biosystems with a group of people that really do what they're doing, it took us about four months or five months with the old technology. With the new technology, they got the sequence in 48 hours. So, four months or five months, 48 hours. So, you got your sequence in 48 hours. You can send it out to everybody and be like, "Here's a sequence, knock your socks off, try to figure out how you're going to attack this thing." 

Now, you got a ton of different companies that are working on it all at the same time. That's very unusual. We really don't have that, all at the same time, trying to defeat this bow. Well, then you've got a company like Moderna, which is RNA, which is basically like a programmable process. So, you're not developing a small molecule, or an antibody, or I have to inject this into an egg. No, no, I'm going to program this thing to the spike protein. And so, I believe within three days or four days, they had the first shots on goal that they wanted to give a try. So, you see how, all of a sudden, with the help of technology, and the advancement of chemistry, and running all these things in parallel, you just shortened the time frame.

What's an example? Maybe GPT-3 is a great example, right? I'm going to write a five-page paper. You write papers. Writing five pages, that doesn't happen in 10 minutes or 15 minutes, right? But I go to GPT-3 and I'm like, "Here, look, here's the subject, here's the background here's whatever, write me the paper or at least write me the outline and give me some major points." It finishes that whole thing in 30 seconds, okay? Now, what would have taken me maybe five hours, six hours, eight hours or maybe a couple of days, I have my first part of it in 30 seconds, and I can start to work with it from there. And so, when you look at what we did on the sequencing side, much faster, on the development of the first targets we're going to throw at this thing, much faster. We've already done the manufacturing facility that's ready to go. Everybody's figured out how we're going to ship this around the world, by the way, not trivial. The files that we're going to need for it, right? 

If you actually go through and listen to the stories of the groups that have been working on this, I think everybody deserves a medal. There were only two companies that make the glass vials that you could put these vaccines in. You had to have them make enough glass vials to get this out there. How much dry ice is out there to ship this stuff around? Do you have enough to ship everything where it used to go? All of these things had to come together for this to happen in that nine-month time frame. I think that's the part that always-- People get caught up in the big, I don't know, some major theory and they're not looking at the details of actually making all of this happen. There was a lot of technology, a lot of people, and a lot of effort to bring it all together.

Melanie Avalon: A lot more context than people probably realize. [laughs] 

Harry Glorikian: Oh, tremendous context, but most people-- Look, most people are just, they don't put the time into really understand. It's the microwave dinner thing, not the, let me cook this thing from scratch view of the world. And so, if you start to dig around and you just say, "Okay, well, this process starts at A and finishes at C." Holy crap, all these pieces have to come together. How do we make them all happen? How do we speed them up? How do we apply technology to get to answers faster and every other piece that needs to happen? Oh, by the way, if I stack these on top of each other, these people have to take unbelievable risk financially to make this happen. Everybody did that to get to the end game. Otherwise, we wouldn't be where we are. 

Melanie Avalon: Yeah, I remember when COVID first started-- Because right when it started-- I was talking about getting flagged with this episode. When it first started, people were talking about it more without concerns of getting censored or anything like that. So, I was like, "I'm going to do some COVID episodes, right now." This was in March. So, I interviewed David Sinclair right at the beginning. We talked about the vaccine and he basically said that, "Yeah, we could have that tomorrow." It's just a matter of basically what you just said, that all of the stuff is in place to get the basic information pretty fast. It's just a matter of then doing the checking, and the testing, and fine tuning. Was it two years later, or a year later, or was it that same year? I don't even remember now when it actually came out, the vaccine. 

Harry Glorikian: We're in 2023, early 2021?

Melanie Avalon: Yeah. Awesome. Well, for listeners, definitely check out The Future You. There's so much in that book that we didn't even remotely touch on. Harry gives all of these examples. Speaking of, like you were saying in the beginning, there's all of these manifestations of AI that people don't even realize are happening, and very cool companies and programs, it's very, very-- I made a list of like, "Oh, I got to get that app. Oh, I got to get that." The skin cancer, I really want the skin cancer one that looks at your skin. It's like, I should do that. 

Harry Glorikian: There're new ones coming out all the time. It really is hard to keep up with. I do always tell people depending on your medical condition, go talk to your doctor and all that. I think that in the next three years to five years, this is going to have such a profound impact on people's lives that they're going to wonder how they lived without it. 

Melanie Avalon: I think so. Well, thank you so much for the work that you're doing. I have three-- They're super rapid fire, super rapid-fire questions. One, you talked about this in the book. So, it is true that, when we're on a website and it's like pick the picture of the car, that that's actually helping train AI? I was wondering if that was true. 

Harry Glorikian: Yeah, it's like the Mechanical Turk thing where you're doing the work [laughs] for someone else to train a system, and you're also convincing the system that you're a human being and not a robot. 

Melanie Avalon: That's cool. Multitasking at its finest. Rapid fire question number two. What's your favorite Star Trek episode from the original series? If you describe the plot, I can probably tell you the title, maybe. 

Harry Glorikian: Oh, I'm not sure about the title. I'm trying to think of, was it the one where they go onto the planet where everybody has Godlike powers and then Bones figures out the formula and gives them Godlike powers. That's a good one. 

Melanie Avalon: Godlike powers. Not the one where they meet the God, the Greek Gods. 

Harry Glorikian: They meet the Greek Gods. But remember, they get thrown around and everything else and then Bones figures out they have this certain molecule in their system, if I remember correctly, and injects it into Kirk and some of the others, and then they end up developing even more powerful. 

Melanie Avalon: If it's the Greek Gods ones, it's Who Mourns for Adonais. I love that episode. But it might be a different one. But the Greek Gods ones. 

Harry Glorikian: But it goes to show you, modern technology, you can solve a lot of problems if you put the right minds and technology against solving that problem. 

Melanie Avalon: I love it, I love it. And the last question is the question that I always end the show with and it's just because I realize more and more each day how important mindset is, so what is something that you're grateful for? 

Harry Glorikian: Oh, I'm grateful for what I get to do every day, honestly, When I think about it, I'm like, "People do other things." I'm sure they love it, but for me personally, I would be bored out of my mind. I get to look at the coolest stuff and play with the coolest stuff that actually makes a difference in people's life. I love it. I'm grateful for that I fell into what I'm doing and I love it. 

Melanie Avalon: I love it so much as well. I feel the same. I'm like the consumer version. I just get to learn about all of this, and then play with these things at home when they reach the consumer manifestation, but I'm just so grateful. It's so wonderful. I wish everybody could be living their passion, because it makes life really wonderful. Well, thank you so much for your time. I really appreciate everything that you're doing. I will be eagerly following your content. How can people best follow your work? 

Harry Glorikian: They can go to my website, which is, www.glorikian.com. And my podcast is there. They can find it on Apple or Spotify or anything else, or links to my books and other things that I've talked about and written about. 

Melanie Avalon: Awesome, awesome. Well, we will put links to all that in the show notes. Thank you so much again. It was incredible to connect with you and have a wonderful rest of your day. 

Harry Glorikian: Thank you. You too. 

Melanie Avalon: Good Bye.

[Transcript provided by SpeechDocs Podcast Transcription]