#231 ChatGPT and AI In Genetics with Daniel Uribe

Have you heard? We are launching a Patreon! For those that are not familiar with Patreon, it’s a platform that allows creators (like podcasters) to offer their followers more content, experiences, and even gifts. We have an exciting list of ideas on what we will offer you listeners, but we want you to make the final decisions on which benefits we end up offering you. So get your votes in through our survey. It only takes 60 seconds to fill out. Thank you SO much!!

In this episode, we explore the intersection of artificial intelligence (AI) and genetics, focusing on the potential impact of AI policies and regulations on the field. Why do our show notes look different this week? We decided to stay on brand for this episode and use AI (shoutout Podium) to write this content. 

We discuss the rise of AI chatbots like ChatGPT and BioGPT and their potential to transform healthcare and the genomics industry by assisting humans in interpreting vast amounts of genetic data. We delve into the importance of using AI-powered tools like ChatGPT to enhance our understanding of genomic data while maintaining privacy, and we discuss The Family Vault, which helps parents maintain ownership and control of their babies' genomic data. We also explore the challenges of implementing AI solutions in the clinical genomics industry, focusing on the importance of data provenance, quality, and privacy. Join us as we examine the future of data management in a hybrid world of centralized and decentralized databases, and learn about the exciting developments in the world of genetics.

Daniel Uribe is the father of a rare disease child. He is an expert in cybersecurity, data laws, non-fungible tokens (NFTs) and genomics. Daniel is the inventor of BioNFTs to tokenize revocable digital rights over human biosamples and derived biodata based on data privacy laws to enable ethical AI training in genomics and healthcare.

Top 10 Keypoints:

1. AI chatbots like ChatGPT and BioGPT are transforming healthcare and genomics by assisting humans in interpreting vast amounts of genetic data.

2. AI policies and regulations, such as the AI Bill of Rights, are being proposed to ensure responsible and ethical AI systems in healthcare and genomics.

3. Data used to train AI must be accurate, properly presented, and free from bias to avoid producing misleading or harmful results.

4. AI-powered tools like ChatGPT can help individuals explore their genomic data while maintaining privacy and control over their genetic information.

5. The Family Vault enables parents to maintain ownership and control of their babies' genomic data while collaborating with researchers in federal programs.

6. Data provenance, quality, and privacy are essential when implementing AI solutions in the clinical genomics industry.

7. A hybrid world of centralized and decentralized databases is emerging for data management in genomics, with a focus on digital hygiene and controlled access to sensitive genetic information.

8. Creating sub-datasets and limiting access to relevant data for specific purposes can help maintain control over genetic data while contributing to research and receiving personalized medical care.

9. Trustable, certified genetic testing is crucial as AI continues to play a larger role in healthcare and genomics.

10. The future of genetics will involve exciting developments in the understanding and interpretation of DNA data, with AI technologies playing a significant role in augmenting human knowledge and capabilities.

Chapter Summaries:

AI in Genetics and Healthcare (16 Minutes)

In this episode, we explore the role of artificial intelligence (AI) in genetics and the potential impact of AI policies and regulations on the field. Our guest, Daniel Uribe, is an expert in cybersecurity, data laws, NFTs, and genomics. He discusses the rise of AI chatbots like ChatGPT and BioGTP and their potential impact on healthcare and the genomics industry. AI has the potential to assist humans in interpreting large amounts of genetic data and augment human knowledge. However, the data used to train AI must be accurate, properly presented, and free from bias.

The Best of Both Worlds (10 Minutes)

In this portion of the conversation, we delve into the potential of AI-powered tools such as Chat GPT to enhance the understanding of our genomic data while maintaining privacy, enabling individuals to make informed decisions about their health. The Family Vault is discussed as a means to help parents maintain ownership and control of their babies' genomic data, while still collaborating with researchers in federal programs. The conversation also touches on the importance of decentralized data storage and privacy-preserving technologies for protecting sensitive genetic information from hackers and unauthorized access'

Centralization vs Decentralization in Digital Data (6 Minutes)

In this part of the discussion, we examine the future of data management in a hybrid world of centralized and decentralized databases, focusing on the importance of digital hygiene and controlled access to sensitive genetic information. By creating sub-datasets and limiting access to only the relevant data for specific purposes, individuals can maintain control over their genetic data while still contributing to research and receiving personalized medical care. We also touch on the potential implications of genetic data on insurance pricing and the importance of trustable, certified genetic testing as AI continues to play a larger role in the field'

Learn more about Genobank.io here. You can also check out this DNA Exchange article that Kira read in preparation for this interview. 

Stay tuned for the next new episode of DNA Today on April 14th, 2023! New episodes are released every Friday. In the meantime, you can binge over 230 other episodes on Apple Podcasts, Spotify, streaming on the website, or any other podcast player by searching, “DNA Today”. Episodes since 2021 are also recorded with video which you can watch on our YouTube channel.  

DNA Today is hosted and produced by Kira Dineen. Our social media lead is Corinne Merlino. Our video lead is Amanda Andreoli. Our Outreach Intern is Sanya Tinaikar. Our Social Media Intern is Kajal Patel. And our Graphic Designer Ashlyn Enokian.

See what else we are up to on Twitter, Instagram, Facebook, YouTube and our website, DNAToday.com. Questions/inquiries can be sent to info@DNAtoday.com. 

Juno Diagnostics has developed the next generation of non-invasive prenatal tests so that all pregnant people can access a higher standard of care. Juno Diagnostics is the only genetic testing company on the market that performs NIPS for common aneuploidies on blood samples from a finger stick instead of a traditional venous blood draw. That means you collect your sample on your own time, at home, and still have NIPS at a CLIA laboratory! Juno’s Hazel™ NIPS screens for common chromosome variations seen in pregnancy, such as Down syndrome or trisomy 21, trisomy 13, and trisomy 18 – in addition to testing for fetal sex. You can order this test yourself, or have your healthcare provider order for you. Head to JunoDx.com and use the code "DNATODAY” for 10% off! Keep your eye out for our full episode interview with experts from Juno Dx including fellow genetic counselor Katie Sagaser and Dr. Allison Rodgers. In the meantime, check out JunoDx.com to learn more about Hazel and their other test Birch (which tests just for the sex of your baby). (Sponsored)

Surely you have heard of whole genome sequencing, but what about rapid and ultra-rapid whole genome sequencing? This is an emerging method of diagnosing genetic conditions for quick management. PerkinElmer Genomics offers this incredibly valuable test, which can be life saving for ill babies and kids. Learn more in our full episode (#226) with PerkinElmer Genomics. You can visit perkinelmergenomics.com for more information. (Sponsored)

As many of you know through podcasting I have become an entrepreneur including consulting for other podcasts. Since I don't have a business degree I have learned a lot through podcasts like Porch Talks. The inspiration to start this show was from the host Melissa Bradley who wanted to inform, instruct, and inspire fellow entrepreneurs, especially in people who identify as women, people of color, immigrants, veterans, people with disabilities, and folks in the LGBTQIA+ community (which drew me in initially). So if you are thinking about starting a business or just love hearing stories about how businesses grow, Porch Talks is for you. (Sponsored)

Transcript

This transcript has been generated by AI and may contain errors. 

0:00:02

Hi. You're listening to DNA today, a multi award winning podcast and radio show where we discover new advances in the world of genetics. From genetic technology like CRISPR to rare diseases to new research. For over a decade, DNA Today has brought you the voices of leaders and genetics in over two hundred episodes. For the past three years, DNA Today has won the People's Choice Best Science and Medicine podcast award. I'm Kira Dineen. I'm a certified genetic counselor. And your host.

0:00:56

There are a few artificial intelligence policies and regulations that could go into effect in this year twenty twenty three. We're gonna be exploring these today with genobank.io’s Daniel Uribe with the focus on how it could affect genetics. Daniel is the father to a child with a rare disease. He's an expert in cybersecurity, data laws, NFTs, and genomics. He also invented IO NFTs to tokenize digital rights over human biosamples and derive bio data based on data privacy laws like GD RP and CCPA to enable ethical IA training in genomics and healthcare. Thank you so much, Daniel, for coming on the show. Thank you very much, Kiran, for having me in your show. This is a a a very exciting day for me. Yeah. And and I don't know too much about kind of all the the AI side of genomics and all the other things I mentioned. So it's a exciting interview for me to actually learn alongside the audience. So I wanted to start because a lot of the talk this year has been with chat, g t sorry, chat, GP, tea. I feel like a couple years from now, people can be like, how did she not know what that was? So can you tell us about your thoughts on the rise of AI chatbots like chat, GPT, and BioGTP, and their impact on healthcare in the genomics industry. Totally, no. Thank you for the opportunity.

0:02:20

So basically, there are many emerging AIs, and I think we will see more coming hopefully. And basically, the the the AI allows the the human knowledge to be augmented. Right? Is is like how an individual can be assisted by these systems and now be able to summarize or read, let's say, peer reviewed papers, like, ten thousand peer reviewed papers. Right? And then have something. So this is going to become very powerful or or is already powerful, but is we we're gonna to see more, not less.

0:03:00

But there are challenges in terms of what data are we using for training these AIs. Right? Is this data properly presented? And is this data really accurate? Right? This is that's the other because this a a famous saying like in AI garbage in, garbage out. So we want not to have that that effect on AI or to avoid it as much as possible. Yes. Yes. It's going to be very important looking at what data we are pulling into this so that we can be getting really good results out as you were mentioning there.

0:03:39

And obviously, with genomics. Big data is a big part of genomics. It's like, you know, how much data we get even from one genome is just astronomical. So looking at even just how we're gonna start using and and even have started using these chatbots and everything is is I think gonna be a game changer just in general, but also within health care and genomics.

0:04:01

What role do you see AI playing in the regulation of healthcare data and privacy in the future? I know privacy is a big focus of yours with unibank dot I o. I mean, how can we ensure that this AI is being used ethically and responsibly? Because, you know, when we start talking about, you know, if we can, should we, that phrase that we say a lot on the show here, you know, that certainly comes up with AI because things kind of start leaving our hands. Not totally. This is this is core for AI, especially for credibility. Right? Because there are some now of or prominent actors in the field that have criticized a little bit chatbot because the the chatbot. Sorry. Because it makes some mistakes. But it's it's it's a very new technology. So it is improving because it improves as we interact and and as it learns and and there's humans feeding with the correct data.

0:05:00

But again, the regulation will play a super important mission or, yeah, the regulation will allow us literally to have, like, trusted AI system versus maybe not trusted or not not authorized. Right? The same as when somebody in in their company said, hey, I am GDPR compliant, or I am CCPA compliant, well, this the exact same thing will apply to AI. So the it will either say, hey, my AI has been reviewed or maybe not. Right? So take it from a granule sale. Either way, Either way, I think in health care and genomics especially, AI will be a great, a great, great, advice or adviser, it'll be some guidance or approximation statistical approximation, but would never be, like, enough to give a diagnostics. Right? So we will still need the experts on each of the field in order to just give a diagnosis or a proper alignment. But again, the AI regulations are coming.

0:06:21

We have two prominence, which is when the European and now President Biden has proposed a bill of rights for AI systems as well. Yeah. So let's get in to that a little bit. I was doing just tad of research just so I knew enough to ask you questions. And I really wanted to learn the moment. So as you mentioned, In the United States, the White House Office of Science and Technology Policy released a they called the blueprint for AI bill of rights, which I think I'm just like, wow, what century are we living in? This is kind of feel feels like a science fiction movie. So what is this bill? What guidance does it provide for genetic companies? Another kind of health care related companies like ours? No, totally.

0:07:08

I mean, both AI frameworks or frameworks to regulate AI, better said. Their their focus on on, you know, inviting the colleagues system to build responsible and ethical AI systems. Right? So it's more into the the guidelines. And recommendations today. But eventually, I think it'll have even some potential even, I don't know, maybe penalties, right, from from some of the authorities. But basically, they are in in in in risk categories. Right? There are four categories at least in the European regulation. There is, like, the minimal risk. And to the to the maximum risk. Right? And it imposes requirements for each category.

0:08:02

So again, we are starting the conversation because, obviously, the governments are worried that the AI systems can discriminate, can decide in a in a biased kind of sense. Right? And and again, we we want to emphasize or these regulation emphasize the importance of transparency, accountability, nondiscrimination, etcetera. And I believe this is this is the correct time to start the conversation about this. Yeah. It really is because as we've seen in so many areas of healthcare and science related to that, we have technologies that are developed before we have laws in place. I mean, we saw this with CRISPR. So I think it's interesting technology moves so fast and laws move so slow where we do have this lag where, you know, companies are able to do things that aren't illegal maybe yet. So I think that's one kind of concern I have, especially in the privacy spaces we were talking.

0:09:07

But kind of looking more at how advantageous this is as a tool. As you said, it's not going to replace a lot of human jobs and tasks and executive functions. But it is a really good tool to be helping us and, you know, I'm thinking interpreting DNA data. That is a huge area where That is what takes the longest amount of time and we're looking for genetic testing results to come back. It's the interpretation part that takes a long time. So having a tool that helps you interpret I can see that being really helpful. How do you see AI helping us to better understand and interpret DNA data and And how do you see this having implication when it comes to personalize or precision medicine?

0:09:48

No, Tully. So humans we are incredibly smart, right, or incredibly slow. Yes. Right. I agree with that. The the computers are the opposite. Right? They are incredibly fast but incredibly dumb in essence. Right? And and hopefully, no no AI is listening So It's okay. We can offend the AI. Like, that's fine. Let's just not offend humans. Exactly.

0:10:15

So the the combination is the powerful one as you were saying. Right? Because again, there is no human kind of capability of even analyzing one genome by hand. Right, let's say, or just by watching the basis, right? So now imagine if you are using or you want to do a population study where you have ten thousand, let's say, newborn sequencing, datasets. Right? So or or whatever. I mean, it would be literally impossible.

0:10:48

So genomics is is is is what it is today. Thanks to computers. Right? Now AI can learn to be even more efficient every single time. Right? And they will become a tool specifically for for for for regional genomics. Right? Or adnicity based genomics. What I'm trying to tell is, like, as you as you know, obviously, the variants of the of humanity is based on on each country, on each ethnicity, like for instance, I come from Mexico, and there are at least sixty eight prominent ethnicities. So we are very diverse country in the United States even more. And such and and so on and so forth.

0:11:33

But but at the other day, this AI is, like, the the the computing capabilities plus the learning. Right? And the learning to automate. This is this could be like the the the main feature. The automation. So that's that's a given. Right? That is happening today. But again, the surroundings, the the governance of these AI's is what is really challenging and it's in every single panel I've been participating for the last four months.

0:12:06

Howard Bauchner: Yeah, it's been really the focus of the conversation. And I'm also thinking about some of the challenges the companies are gonna face in terms of actually implementing these AI solutions. We're talking about how great it's going to be, but we have to get there. We have to be able to learn I use these tools? You know, especially in as we've been talking about with, like, the clinical genomics industry, what are some of these challenges that you see when it comes to implementing, to start using more of these AI technologies within data and everything. Well, that's that's the key question I would say. Because today, I mean, it's it's a saying. So cyber security, and the the basic of the the encryption and the encryption are given today. Right? So you that that we have to put that like the first rock. Right?

0:12:58

But the the big the biggest challenge today is the provenance of the data. Right? The provenance of the data or the, let's say, the certification of data sets that are conflict free that are consented, that have a proof of integrity, have proof of quality, have been authorized maybe by an IRB, by an independent research board. I mean, that's the challenging part. That that is what we call the metadata of all the genomics and health care data. Sense. Right? That's that's the important part in my opinion right now. So it's not only if I have ten thousand high quality data sets is what is the the the legal framework, right? What is the the the authorization, the the permissions that I have to use the data. And that is going to be the first question when an AI shows like, it's it it is accurate. Right? Or is predicting accurate? That that I'm I'm sure the first question would be, is this working? Yes. And later on, right on would say, okay, let me show the credentials or the certifications. Of the data that you use to train your AI. Because otherwise, it will be a potential liability right there. Yeah. Yeah. Certainly important so that we see well, how is it coming up with this information? Like, what are you feeding into it? Like you said at the top of the show? And and part of what Genobank dot I o is looking at is a lot of this. So you guys have developed really cool technology especially to make sure that data is safe. So can you kind of further explain how genome bank dot I o's tech actually stacks up to help individuals and families explore their DNA using AI while also ensuring that that data is safe and secure. Of course, no. Thank you.

0:15:01

Aveli, this is one of our main focus right now. General Bank started in two thousand eighteen with one main focus. Like, let's enable patients like individuals and family to own and control their DNA data set. So it's just like a genetic data locker, right, where you can interact with the laboratory, send a sample, they will upload your data in a private repository. That that was pretty much it. Then we we expanded the functionality. So now you can connect it to bioinformatics. Right? So now if you want to, let's say, compute your ancestry or your admixture, you can do that inside your genetic data locker with privacy.

0:15:47

But now with these emerging AIs, so we just connected to an AI. So you can now interact with chat GPT. And it's very interesting because chat GPT, you might ask, hey, chat GPT, please. Least the snips that are associated with Alzheimer's disease. And it'll it'll tell you the list. And now through our our system, you can literally go through through some code and list these snips and then make a compute over your own data with privacy. So this is done without disclosing your private to any company, not even to us because the the the compute is done in a private repository.

0:16:38

But again, I think this is or we think this is the correct approach, is not necessarily connecting the data directly to chat GPT or to bio GPT or whatever the AI might be, but is to just combine the two of them and respect again the privacy because you don't you don't want to necessarily disclose your risk to develop Alzheimer's, right, right away. And and even less to an AI that maybe or maybe not let's say, like for instance, Chat GDP has a huge investment from Microsoft. Nothing against Microsoft, but maybe I don't feel comfortable disclosing my genetic information to a chatbot that is partially owned by Microsoft. Right? That would be the case. But again, I want to to have the the ability. I want to increase my superpowers human powers in terms of using these these AIs to make an approximation to what what is what is going on with my snips, right, with my genetic data. But again, this is just like to the first thing form. That the best thing I I I would say is the interactions later on. So when you can book, maybe a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a and say, hey, you know what? This is a report that my genetic data combined with their common of chat GPT or bio GPT throw. This is some references. Please help me. Right? And they will literally be able to help you. But the end of the day, this is again generating more and more conversations. We call these citizens genomics.

0:18:30

Is not necessarily that as a city centers or as a patient or potential patient, you will go directly to the diagnose. No. There will never be the purpose. But it's again it's like the the the we come to the from the era of of doctor Google. Right? So you have some symptoms, you will ask. So now if you have your genetic data, you will ask chat t p t, say, hey, how do you reference, you know, the the snips for developing atherosclerosis. Right? And you can do that in our platform and give a statistical report without any privacy concerns. And it's so helpful for patients to be able to take this power into their own hands and say I can do research and I can collect information. And then when they do get to meet with a healthcare provider like a genetic counselor, they can have very pointed questions because they can be educated to that point. So I think that's just huge because a big, you know, goal of mine is is to be able to contribute to these efforts where we are developing tools and helping people educate themselves so that then when they meet with these healthcare providers, they're not like learning, well, okay, well, what do you mean? What is a genetic variant? Like, they they can learn that off something else, but really getting into these conversations like you said. So it's it's such a great marriage of the two of of using the power of these tools, but in this setting where it is private and safe. So it it really is the best of both worlds there and And the other thing I wanted to ask you about is your family vault.

0:20:00

And and how that helps parents participating in newborn sequencing programs like, you know, baby seek. I mean, there's baby deer, baby bear out there that are doing, like, a whole genome sequencing, rapid whole genome sequencing. And this is area that we've talked about multiple times in the show. So tell me how this helps parents maintain ownership and control of their babies genomic data, but then also be collaborating with researchers in these federal programs to help, especially in the rare disease space, because I know that's one that's close to your heart. Howard Bauchner: Exactly.

0:20:35

And this is and thank you for your opportunity to talk because we I mean, we got diagnosed. Our our kid was diagnosed until he was two years old. Right? Which is timely. I mean, there's some other not so lucky brands that even take longer, like even five years. To eight years I've heard most recently. It's like an African diagnostic odyssey. It's so long. Exactly. So that's one of the key aspects. So imagine when you have your baby, then weeks later you have this report and you have actionable item. Right? This is even if this is not a a life threatening situation, even now, like Stanford, doctor Elon has and his team has the record. They they they had the newborn sequencing data at results. Yeah. And he was on episode one fifty, I think. Exactly. And he he chatted about that about that. Yeah. Like, amazing. It is amazing. So we are there. Right?

0:21:34

So again, the the big conversation right now is who is the correct custodian? Who is the owner? Who who is is the one that will give access for how long. Right? And such and such. In our point of view, again, the family board is inspired directly in this So it's a genetic data locker, but now it's not for the individual, it's for the newborn, but also can get oily mom and dad and the siblings inside it. They ecos genomics is a family reference, right, at the end of the day. So it can be better studied if you have obviously the the the the complete family. Or as much of the family members as possible.

0:22:19

That being said, the main important thing is how you can collaborate or still collaborate with these very important research programs that we have now available. And we also protect them because these companies are getting hacked. Right? So just a couple of days ago, one company in the UK was fact, two point one million genetic datasets were stolen. Right? And this is really, really bad, and we don't want that to be the case of of the baby of the babies of or the newborns. Right? So again, we are using decentralized born. So instead of having just one data set database with millions of records, you will have each of the data set separated, properly separated, right, in the governance of each family board.

0:23:11

That's our vision. I don't know if this is going to be really the future, but this is should be ambition. Right? So it's literally the hackers will literally have to hack each and every one of the boards. That elevates exponentially. The amount of resources, again, negative, focused, but the amount of service necessary to decrypt every single bolt. Right? And And also now if if that those bolts can be connected to AI and somehow ask some questions with privacy preserving technologies, as sero knowledge proofs or homomorphic encryption and such and such. Then we think it's a very powerful tool and hopefully then population genomics at the level we need of humanity can be developed.

0:24:02

Because if somebody is without the respect because I've been hearing some people talking, like, if somebody expects that there's going to be a database of, let's say, three hundred and fifty million American against all aggregating one, open protected database. I think that's even far from reality. That's something not not gonna happen. Yeah. Hopefully, it's not gonna happen. Right. Hopefully, not gonna happen. Because if that was there, then that's like easy target at that point for a hacker to be at least trying to have that be their goal to get into that data and and steal that data.

0:24:41

So it's it's interesting, especially when you're looking at it from perspective of different countries government. So in other countries, we have outside the US, like in Europe, like there's the healthcare is much more It's simplified. It's through the government. Whereas in the United States, it's through a lot of private institutions. So I can see how that there's differences there and that other countries might be doing a lot more gathering of genetic data and having in one database, which I would think is as we've been talking about a little bit more susceptible to hackers. Whereas what you're talking about is kind of having a bunch of different ones. How do you foresee with your company with Genobank dot I o? How do you foresee your bio data helping governments with these solutions and other, you know, in the US, like small maybe not even small, but like someone like, you know, Yale or Mont if you're, like, these larger institutions, you're a Presbyterian where they have their own databases, but they're still really big. So how do you see you helping kind of really shape the landscape of genomics and making sure this data is accessible to who it needs to be, but not accessible to other people.

0:26:03

No. That's that's I believe that that's a core question. So we're gonna leave in a, I think, in a hybrid world for a while. Right? Centralization and decentralization in terms of who holds the data, like centralized database as versus decentralized databases, which the decentralized basis is what we propose in the family board. It's basically a database in a family owned server or instance or machine, right, or or cloud, whatever that that may be. Versus, as you were saying, like, a government, like, centralizing all the information. So the the only thing that we we think is that Nobody should have your whole genome, for instance, right, or your whole exome.

0:26:50

So we are thinking more into how can we increase the digital hygiene, right, or be more hygienic with our digital data. So this is some something, like, is when you use the services from a hospital. So so when you go, you you use the services, but you keep your stuff with you. That that's more what the the world we envision like, hey, you can port I mean, in a portable kind of state of the data, you will be able to open access, cloud access, choose certain regions of your genome. Like, for instance, just create sub datasets. Like, for instance, in oncogenomics. Right? We know, like, for cancer genomics, breast cancer genomics is approximately three hundred genes. But not the twenty one thousand or the twenty two styles. That was an ish. Yeah. Whatever that is. They don't need access to all that. Just the ones relevant to cancel. So why don't we just create date sweep datasets that is on clicks on your computer and say, okay, these are my oncogenes. Right? And this is enough to whatever the government wants. Right? So that's kind of a better approach.

0:28:07

Obviously, the the the vision or the dream is that everything is encrypted using homomorphic encryption and such as but but the amount of computational power to deliver homomorphic encryption in just one human data set is huge. Today, is is not not even a question. Right? I mean, it'll cost like thousands of dollars to make just one query of of, I don't know, few few data sets. But what I'm trying to tell you is, like, it's it's going to be a negotiation. Right? Because it's true that the researchers love to have the data with them. Right? Available in their servers, in their premises because sometimes they have more powerful machines and so on and so forth. But let's not forget that now in in our hands, we have very powerful phones in in laptops that we have the m one, the m two chipsets from I mean, these are literally AI prepared laptops. Like, they can do a lot of the computing that we have dream, like, ten years ago to to do it in a very large computer in a data center.

0:29:21

But but again, I mean, the the the things changes, but I think we're going to live into a hybrid world for a while. Yeah, I think we are and a lot of this is important. To be able to keep safe and where you only have access to it or people have only limited access as you've been talking about because insurance companies, that's a whole aspect that gina protects us from some things, but it doesn't protect us from everything in terms of insurance companies saying, well, we're gonna charge you a lot more because you have this variant or something. So it doesn't protect against life insurance, long term care. So I think that's something that maybe even that stuff will change too, where suddenly that's going to be included in the future where we are protected, where insurance companies can't hike up pricing based on your epidic results from that. So I just wanted to mention that at some point during the episode.

0:30:10

But Daniel, thank you so much. I mean, you really know your stuff. I just learned I knew I'd learned so much in this episode, and this is just gonna become more and more relevant as people are doing more genetic testing especially as you were bringing up in terms of like in the NICU and even doing like newborn screening where eventually I think because we've been predicting on the show, it's gonna start being whole exome sequencing at some point, I think, with newborn screening. So So Yeah. We're gonna need you, so please stay in the field. Thank you. No.

0:30:43

Hopefully, we will we will be able to serve as many programs, individuals, families and such and such because at the end they we were talking. Right? We we need to to we have genetic testing. Thank God for genetic testing, but we need certified and trustable genetic testing. Like, hey. Sort because we we are going to now to connect it to AI. And again, AI will amplify the errors or will help us to amplify the the correct data. So we weren't always the correct data. We want to use the the tools and and and we need the the certification of data for this case. And hopefully, we can help everyone that is interested in this type of solutions. Yeah. Definitely. Well, thank you again. Thank you very much, Kira. I appreciate it's this is a show that I love, so I'll I'll I'm very honored hear it with your audience.

0:31:46

For more information about today's episode, visit dnatoday dot com or you can also stream all two hundred plus episodes of the show, including video versions of interviews recorded in twenty twenty one or later. Any questions at episode ideas, guest pitches or sponsor inquiries can be sent into info@DNApodcast dot com. Be sure to follow us on social media, especially so you don't miss a giveaway. We are at DNA Today Podcast on Twitter, Instagram, LinkedIn, Facebook, and more. Please rate and review the podcast on Apple's spot phi or wherever you listen, this truly helps us climb the charts and allow more genetic nerds like yourself to find the show. Here's a bonus. If you do and send us screenshot, I'll give you a shout out on the show.