December 2023
This is because organizations may experience initial productivity dips and will have to teach their workforce new skills when integrating AI. AI will ultimately produce more efficient systems for most companies, but the initial cost of time and money may discourage them from adopting AI as quickly as individuals. Mike and Sasan agree healthcare and education will be heavily improved by AI. Artificial intelligence can help enhance curriculums, provide data analytics for educators, and improve accessibility for disabled/neurodivergent students. In the health industry, we may be able to bring AI-powered tools to countries where medical facilities are limited, use AI to accelerate drug discovery, diagnoses, genomic sciences, and provide telehealth services. Some risks AI may present are job displacement, bias in data, privacy/security, ethical dilemmas, regulations, and more. As AI continues to develop, we must ensure it is used responsibly. Looking to watch some of the live sessions from our Investor Symposium? Stay tuned as we release more episodes like these in the coming weeks such as Generating Attractive Returns Through Creative Private Lending and Bond Alternative: A Smarter Way to Lend to Corporate America.
Watch previous episodes of THE FINANCIAL COMMUTE here:
Ep. 62 Lending to Out-of-Favor Companies with Substantial Assets
Ep. 61 Decoding Bitcoin: The Future of Crypto
Good morning. Good afternoon, everyone. Thanks for being here. Great crowd. So very excited to be here. I'm Sasan Faiz Managing Director of Investments, along with Mike Rudow, Wealth Advisor at Morton Wealth. So we're going to be talking about artificial intelligence, which has been a hot topic all year. And Mike and I told you, we're going to do a session on AI.
We asked that our session not to be scheduled right before lunch and not right after lunch. But apparently we just got scheduled right at lunch. So hopefully it's interesting enough to keep you all engaged. We're going to start by talking a little bit about AI, its background. Define some terms in AI. Then we're going to move on to adoption phase of AI for organizations and companies, what challenges they face.
We're going to then go into the really more most important part of the presentation, some of the real life implications of AI and how this is actually affecting our lives and the society in general. We'll have a brief discussion on risks and regulation, and then we'll finish it up with some concluding remarks. And we're going to do that all in 20 minutes.
So with that, I'm going to hand it over to Mike.
Now, thanks Sasan and I'm excited to be on stage with you. I think this is going to be a fun day. I wanted to start with a little bit of history and kind of give a definition of AI before we move forward. So when you think artificial intelligence, a high level definition are machines and computers that are made to think and learn like humans and then complete a task.
What we're going to be more talking about today is generative AI and generative AI uses AI systems to generate original content. So this content like images, sounds, music, movies, and this content is created through natural language. So what do I mean by that? Well, an example would be, you know, I love to read my kids bedtime stories and the same 20 to 30 bedtime stories get old and I lose engagement for my children.
But I figured out a little hack for life and I could go on ChatGPT and I can type in ‘write me a children's bedtime story that's focused on the main character named Ryder who's a young boy. I wanted to involve robots. I wanted to have ghosts’. And, you know, I wanted to have a couple of the perpetual characters.
And lo and behold, it'll give me a five minute story that I could then read to my son and keep him engaged because it's things that he's interested in. And the stories are actually good. So aside from that, let's go a little bit into the history of A.I. which has been around for decades.
Back in the 1980s, IBM started a program with Carnegie Mellon, and that program was to develop a system called Deep Blue, which was a supercomputer that was made for chess. Fast forward to 1997, had uploaded a tremendous amount of data that that supercomputer was able to process that data to make decisions. And the computer was actually able to beat the world champion in chess named Kasparov.
And that was a huge milestone. And if you look at that next evolution of what I can do today, instead of having to upload massive amounts of data to where each move has an implication, all you need to do for current A.I. is give them the rules of the game. If you type in the rules of chess in modern A.I., it'll be able to run its own simulations and then probably beat whoever the current reigning champion of chess is.
So it's come a long way. Another way that you could look at the history of A.I. is just through virtual assistants. Back in the nineties, we had chat bots where you would go on a website and a little box would come up and it would say, you know, how can I help you? And you say, I want to know the store hours.
And it could provide basic information, but you had to type in the question correctly for it to give you the answer. And if you didn't, then you wouldn't get the answer that you were intending to get. That evolved back then to thousands to basic voice commands, where we remember the original Siri or Microsoft Cortana, where you could set an alarm or send a simple text or do basic tasks and now look to the 2010s and current day that has been revolutionized completely to where it's integrated into our daily life at a very high level.
We have our smart speakers that we use to ask questions. We have smart homes where we can adjust thermostats and we can adjust lighting and locked doors at certain times based on habits, based on habitual patterns. And so when you look at the history of where we've come from, where it is, we see as it starts to make lives a lot simpler, a lot easier, we can be a lot more regimented with our time and focus, our energy and on things that are maybe more important.
And it's going to go along the same for businesses, right? Businesses are going to have a huge tailwind with the implementation of AI, and we've seen the adoption of that on the customer side pretty quickly. Right. There's been an unprecedented level of adoption when it comes to individuals. I don't know if you guys are familiar, but it took Netflix three and a half years to get a million users, which was a pretty astounding benchmark.
It took Facebook ten months, it took Instagram three months, it took ChatGPT, five days to get a million users, five days. So from an implementation and adoption side, it looks like we've all bought into it. But from an enterprise level size from the big businesses, I don't think that adoptions going to be as easy. What do you think?
So do you think there's going to be challenges and headwinds to adoption?
Yeah, absolutely. I think that A.I. is going to be a transformative force in the society, in the global economies. But adoption for organizations and companies is going to come with huge challenges. So most of you invested in maybe private equity type closed end funds are familiar with the concept of J curve.
So that means that initially a strategy is going to have a dip in returns because of fees and expenses. And then as assets are added and they generate return, their returns move into the positive territory. The same concept applies to adoption with organizations as well. So there is an initial disruption where productivity is going to dip. So that's associated with some of the traditional workflows and processes are going to become obsolete or they're going to have to be modified.
And that is by the time the companies get going, kind of recognizing what the issues are and how to how to incorporate new processes, there's going to be an initial dip in productivity. The second stage, Mike, that we've identified is is kind of the reskilling and adaptation that has to go on with the workforce. So the workforce has to be trained or retrained to be able to work with an AI system and be able to use it effectively.
So that's also going to take time. It's obviously dependent on each industry and the readiness of each industry to be able to adopt artificial intelligence. The third stage is really the where some benefits of AI is going to materialize. So we are looking at companies at that time. Once these initial hurdles have been overcome, they're able to optimize their processes and they're going to be able to use AI to basically improve productivity at their organizations.
The last stage is really the enhanced productivity that's going to come in once AI is fully deployed across different industries. I was just actually reading a few weeks ago a report from Goldman Sachs, and it's talking about productivity is going to grow at plus one and a half percent rate once AI is fully deployed. So when you think about the growth in the economy, there are two drivers.
Mainly the working force population growth and also the productivity growth. So working force population growth has been a little bit stagnant in the U.S. So the majority of the growth in the economy has to come from productivity. And if Goldman is right about the projections on productivity, then it's going to be a huge game changer going forward. So kind of the kind of shifting toward the some real life implications of AI,
Mike, you and I have kind of identified two areas in global health and education where there is huge need but not enough human capital.
So we think that those are two areas that are going to get impacted the most. If you can maybe start by talking about how AI can get integrated into education and then I'll move on to health care.
Yeah. Thank you. I think education is one of those topics that impacts all of us, whether it's for ourselves, personal education, for our kids, for our grandkids. We can relate to our own struggles, going through our own education processes and experiences. But AI is going to change the education landscape tremendously. And I think the three biggest benefits that we can see before we dive into example are first personalization.
When you have an AI system implemented in a school, it'll really help the kids quickly identify their strengths, weaknesses and their learning type, right? And that way it's not up to the teacher to have to focus individually on each one and try to create a specific learning path. I think second is data driven decision making. Having the analytics to make decisions on problem solving, solving problems or resource allocation and creating strategies for troubled students is a big deal.
And then the last, I think, big improvement that we'll see with education is accessibility and equality, right? With the new type of systems, you're going to have the ability to level the playing field for disabled or for people with geographic or financial restrictions because of things like speech recognition or speech to text or translation and 24 seven assistance.
And I think that's the most exciting part. And to kind of bring it all together, I wanted to give you a real life example. I have a daughter. She's five. But for the story, I want to pretend that she's a sophomore in high school. So Taylor walks in to her first day as a sophomore in high school, and she sits in her math class and she's notoriously had a little bit of trouble in math, unlike her dad, who is a perfect A student.
I hope you guys are verifying this, but she sits down and the first thing she does to her surprise is take a diagnostic test, an assessment test, because her school had just adopted and I tutor a program. So she goes through this test, and the test then identifies her strengths, weaknesses in her learning style. And it it comes out to show that she actually excels in geometry but is having trouble for keeping up with algebraic equations.
So the next thing that happens is it'll create a customized path for Taylor moving forward. So instead of the next assignment being general math equations, it's going to be have specific algebraic equations based on her level that will start to get harder as she starts to improve. Now with that also comes real time feedback. If she's in the middle of an equation and having problems, it will give her hints to help her and it understands her learning style so it knows how to kind of launch those hands.
If she gets a question wrong, it'll show her why she got it wrong. How she got it wrong, how to fix it. And then the tracking on that is incredible. You've got a dashboard that she can go back to and see where she has come to from where she started so that the teacher could then sit down with her and say, okay, you're really efficient in this area.
You're still lacking in this area. Let's create a personalized plan to where I'm spending time with you to really help you overcome the challenges that you're having with this program. There's also a gamification part of it where it's keeping her motivated because she's getting coins and rewards for the more work that she does. And that might lead to a free homework pass or, you know, a piece of candy which might not work in high school, but it does now.
And so the way that it can keep children engaged and motivated to learn and level the playing field there, I think is is pretty incredible. And then the accessibility. When Taylor was graduating preschool last year, she sat on stage like this and she's what do you want to be when you grow up? And she said, I want to be a hip hop party girl because it was one of my proudest moments.
But she's now going to have the opportunity to be that hip hop party girl. She can go to her hip hop classes after school if she needs to do her tutoring before school, during lunch, after school. There's no excuse she'll have access to that tutoring at any time. So I think
there's a lot that can be done with with education.
And I know you're passionate about medicine. Where do we see the impact in medicine or healthcare?
Absolutely. So I think that the biggest impact right away that I can have on global health care is really going to be felt in the emerging economies or frontier economies. So this is where access to health care is very limited and the majority of under five year old deaths occur. So this is where I can make a huge impact.
So right now, for example, there are ultrasound machines that can be operated using AI with minimal training, and that's one example of improving healthcare in those in poor countries. Obviously, AI systems have to be trained on diseases in developing countries versus developed countries, and there are many restrictions in in developing countries that have to be accounted for. And the AI systems are prone to making mistakes, but so are humans.
So basically it's better to have some access to healthcare rather than not having access at all.
And so kind of moving to developed nations in the U.S., for example, the most immediate impact that's already being felt is really on the administrative side that we can health help health care professionals find insurance claims, organize patient data and document things.
And those are really the immediate improvements that we can make in productivity in the in the health care system. So a couple of examples. Just given that area has, you know, medical records are, for example, they're very dense and complicated and they're usually stored on different platforms. So a company called Meditech right now is using a AI system to organize this data and put it all together under one roof.
So that's very important. There's another company, HCA Healthcare right now is piloting a program where physicians can use hands free devices, basically to securely create medical notes after patient meetings. So once these nodes are created, the physician can review it and it's all going to be stored in electronic health records. So under one roof, again, it's very important.
The main thing in medicine, I think that's going to be a game changer is that I'm personally very, very passionate about is really the convergence of big data analytics and genomic sciences. So I think we all have experienced loved ones dying from cancer and other diseases and there are ways to improve and basically cure these diseases. And I think what's been missing is the technological aspect of it.
So when you kind of looking at biological the records, they're obviously very dense and very complicated A.I. systems can look at the biological systems and come up with ways where they can find drugs that are going to cure diseases. So when you're looking at what modern medicine tries to accomplish, basically they're looking at dysfunctional proteins and they try to find another protein that's going to basically modify the behavior of that dysfunctional protein.
And that's what's called the drug. That's how a drug is developed. So when you're looking at proteins, there are complicated molecules. They're made up of over 100,000 amino acids. So they're very complicated to analyze. And AI systems can help to do that. Right now, when you're looking at kind of creating a protein that's going to dock to another protein to basically solve and cure a disease.
There is a lab at the University of Washington on Baker Lab that's using an NVIDIA chip basically to design a 100 amino acid protein in 11 seconds. And now when you look at what used to be done, it used to take eight and a half minutes. So that's almost a 50 x improvement in being able to create a protein that can dock to another protein and and cure these diseases.
So I'm very passionate about this area of medicine. I think we are maybe at the beginning stages of a golden age of genomics.
So again, we've talked about some of the benefits, like why don't you maybe go through some regulations and risks that you see in AI?
Yeah, I think that's probably just that's top of mind for people. We are running out of time, so I'll try to keep this brief. When you think about regulation as a headwind, the US takes leadership in technology and AI very seriously. They see it as a true necessity for national security. And with the way that our relationship with China has been going, kind of decoupling the tech sectors, we see that as as a real headwind for chipmaking companies and for the next gen companies in technology.
So it's something for us to keep a look out for and for us to be educated on. But then when you look at other risks, I think what's top of mind for a lot of people are job displacement, right? Is A.I. going to come in and replace a jobs? And it's very possible that for task oriented jobs that are simple and repetitive, we're already starting to see that in things like fast food restaurants.
But as opposed to being scared of job replacement, I think we need to refocus on what the skill set needs to be for the new jobs that are going to be created right now and making sure that we're equipped for that moving forward. Another risk is bias. And that is a big risk with AI because the output is only good as the input.
And for a lot of data that's out there, there's inherent bias. And if there's bias going in, then the risk is that there's an accelerated bias going out, right? So I think whether it's in things like lending or criminal justice or hiring, we need to be really careful and ethical on the data that we have that we're using.
And then the third risk I think that's top of mind for people is we are at 20 minutes, so I apologize and that is that I talk too much. The third risk is privacy and security. And with that, I think we've all started to experience that. I think you read stories about a grandfather who gets a call from their grandson and he needs money because he's in trouble.
And lo and behold, he provides the money and he finds out that it wasn't his grandson, that there was a program that took his grandson's voice. Right. So I think security is a huge topic that that is a risk moving forward and also something that that needs to be educated on. But if you wanted to dive into the winners and losers, I think we are over time. But Mike and I are going to be around, I think, to discuss if you have any questions. We talk about winners and losers. So again, our presentation was on A.I. Artificial Intelligence, and I promise the rest of the sessions are all about real intelligence. So thank you very much for your time.
Thank you, guys.
Disclosure: Informationpresented herein is for discussion and illustrative purposes only. The viewsand opinions expressed by the speakers are as of the date of the recording andare subject to change. These views should not be relied on as financial, tax orlegal advice.
