Conversations on Applied AI

Jon Herke - AI & The Next Generation of Emerging Technologists

• Justin Grammens • Season 1 • Episode 3

Share episode

I'm thrilled to have had the opportunity to have this awesome conversation with Jon Herke, a leader, and technical evangelist based in the Twin Cities. In this episode, we cover a TON of ground on everything from how Emerging Technologies such as Artificial Intelligence is being taught in schools and how Jon is intimately involved in fixing that through the Futurist Academy, to what's ahead with AI and Graph Databases and the J.A.R.V.I.S project and the Neuralink Corporation, founded by Elon Musk. I hope you enjoy this fascinating conversation with one of the most humble and giving person I know. As he shared with me, the three words that he lives by are: "Learn. Create. Share".

Finally, if you are interested in learning about applying AI, be sure to join us at a future Applied AI Monthly meetup and help support us so we can make future Emerging Technologies North non-profit events!

Here's just a small sample of the topics we discussed. Enjoy!


Jon Herke :

really believe in the philosophy of giving value to the world, I don't know where it's gonna lead you. But if you just put yourself out there, continually producing value, things will just come upon you learn these technologies. I've learned how to build a blockchain on the theory and test net. I go to a high school and I would give a guest talk and I would convert the whole classroom into a blockchain and we would actually execute a whole consensus and the students were actually doing the transactions and being the blockchain so that they had a solve a mathematical equation.

AI Announcer :

Welcome to the conversations on applied AI podcast where Justin Grammens and the team at emerging technologies North talk with experts in the fields of artificial intelligence and deep learning. In each episode, we cut through the hype and dive into how these technologies are being applied to real world problems today. We hope that you find this episode educational and applicable to your industry and connect with us to learn more about our organization at applied ai.mn Enjoy.

Justin Grammens :

Welcome everyone to the conversations on applied AI podcast. Today we have john herkie, developer and technical evangelist at Tiger graph and also founder of futurist Academy. Tiger graph is scalable graph database for the enterprise. And with their proven technology, it helps to connect data silos for deeper, wider and operational analytics at scale. futurist Academy is working to reshape the future, they leverage the latest technology and innovative ideas to mold a future of self empowerment with their platform, educators will be able to host their creative teaching solutions, styles and agendas. In return, students will be able to pick which courses best fit their learning styles or need besides all of this, which is which was a mouthful, John's got a very interesting background and career. He's worked for everything from fortune 100 to startups, and most importantly, john, I don't think I've ever told you this. But I appreciate and thank you for your service to our country as a member of the Army National Guard. So we're here today to talk about AI. And I gave a little bit of a background I guess, where you're currently at but you know, if you could maybe share with the audience a little bit about Yeah, what you've been doing here and your career?

Jon Herke :

Yeah. Yeah, career. It's an interesting, interesting word. You know, you grow up and you think that you're going to be something and then, you know, life just comes at you and and you're always not knowing what's going to be in the future. And I tend to live in the moment and I don't know where it's gonna go. But I'm excited to be here. It led me to you led me to all the great peers I have, and our home state here in Minneapolis. So a little bit about me. My background in the military was a computer networking engineer. What that meant is that I would go out in the middle of nowhere, shoot up to a satellite, grab an uplink deploy network. So just imagine a barren field of nothing. And then you bring all these big boxes in and it has the networking equipment, routers, switches, and you bring big trailers that have these dishes that shoot up to the satellite, and then you connect everything, program everything. So that's what I did in the military.

Justin Grammens :

Were you always into technology, kind of a geek growing up at all?

Jon Herke :

I would say so. I mean, I was one of those kids. That, you know, built the computers and websites back in the day. You know, back in MySpace I date myself to the younger folks for the older folks are like yeah, you still Yeah.

Justin Grammens :

And I do mean I do mean geek in the nicest of ways actually. right it's it's a it's a loving term. I call myself that all the time.

Jon Herke :

Yeah. Pure technologists love future thoughts and related them back into reality. I did have a start up. I'll we can talk about that later.

Justin Grammens :

Cool. So after the military, yeah, you're out here doing networking stuff, crazy stuff in the middle of nowhere. What kind of led you then to the next phase of your career?

Jon Herke :

Um, yeah. So I network, an engineer, I had a secret security clearance by the Department of Defense, all my certifications. And then I applied for an internship at Uhg. They're like, Man, this guy's like a super intern. You know, he's got six years of experience networking, and he's got all the certifications. And they let me do a lot of cool stuff. They're working for a large company and doing networking a little bit different than being out in the middle of a field. Got a little bit repetitive, you know, you can't go wild wild west on your network. So, so Like, yeah, I got a little I got a little stale I at that point decided to explore different options. And just by happenstance, I got an email that was somebody looking for somebody in a group called the garage in the garage, which was a healthcare startup incubator. And so I said, you know, why not? You know, I love entrepreneurship. I love technology. I love being creative and building things. So I applied there and got the position as sort of an entrepreneur in residence. So building healthcare startups. So it's like a 360, from being a very technical guy to building healthcare startups. But when you're building a startup, there's a lot of system architecture, a lot of thought process that goes into that, and deep thinking and execution, so it wasn't totally something out of the blue. But yeah, I did that for a couple years, two different startups. One was really successful. One was not so successful, but I got to learn a lot. And it wasn't just me as as actually working with other founders and stuff there. And then, in August 2016, we spun off and created this group, we focused on emerging technologies called the advanced technology collaborative. The startup incubator was very business model focus. So what that means is if you were to create Uber, the technology existed, people existed, the cars existed, the phones existed, everything existed, it was just a way to rethink about how business would operate. And so that we did a methodology called running lean. That was very small focus. But the thing that we kept running up upon was that there would be a lot of these emerging texts coming out like AI and, and blockchain things and NLP and computers, you know, all these crazy cool things. And we're like, what are we doing to address this space? So back in 2016, we created this division that was explicitly looking at emerging technologies and looking and understanding them and how they could fit in the healthcare space. So a good example that everyone probably could relate to, is if the internet came out, what is it? Why do we care? Is it a fad? And is it gonna be here next year? Sure. Can we make money off of it? There's so many questions, you know, oh, yeah, that's that important, you know, the internet, like who's going to use that, you know, there So many questions and unknown variables when when new technologies come out, and people just don't understand them. So what we wanted to do was hire people that truly understood these technologies, and that we had the people that truly understood the business, and then put these people together. And we built cool things. Nice. So that that was my time at Uhg go into entrepreneurship, building out this, this department and then going back into technology. So I really pushed down on graph technology. So we worked on blockchain, Ai, graph databases, quantum computing, these are just general emerging technologies. I emphasized my passion and networking into graph did that for a few years. And then I got picked up by the company Tiger graph. They saw what I was doing on the community doing things with futures Academy. Yeah, well, that's Yeah, let's let's talk about futures Academy. So that's already good. I mean, you've built up this deep knowledge, obviously, working with people and all these really cool emerging technologies. One of the best things I think about your story john has just really sent you love to get back I really, really love being in community. So it should be a good good time, I think just to maybe talk a little bit about what futures Academy is. So the futures Academy really spun up on the knowledge I was accumulating the connections as accumulating the knowledge as accumulating. And as you said, giving back I really believe in the philosophy of, of just producing value to the world, giving value to the world, I don't know where it's gonna lead you. But if you just put yourself out there, continually producing value, things will just come upon you as such meeting, meeting you guys meeting, Tiger graph and everything else. So with that being said, I would learn these technologies. I've learned how to build a blockchain on a theorem test net. I go to a high school and I would give a guest talk and I would convert the whole classroom into a blockchain and we would actually execute a whole consensus and the students were actually doing the transactions and being the blockchain. So they they had to solve a mathematical equation. And then so yeah, so we walk through sort of the history of why electronic currency exists. Why There's this thing called value exchange, the whole reason there's money and then we walked through blockchain, we actually launched our own coin in five minutes on that tyrian test net. So everyone got that experience of understanding that technology. And this is this is a high school class. So you have to really, like articulate and understand the technology and then be able to bring that into a classroom. So I did this. I did that also, along with other technologies, including AI. So computer vision, since you guys talked to Dan McCreary, the AI Racing League, and you have a 11 year olds building self driving cars, I think it's just really cool to say that. And, you know, capturing data, creating models and then taking those models, putting them back on the car and letting them drive around. But the whole point is taking these emerging technologies. I really loved going to different meetups, different schools, and I got the question, how can we be part of this? How can we be part of this all the time? How can we be part of this so then we decided to create an organization around it called futurist Academy who were our primary directive is really looking at these emergent technologies and bringing them into the education space. So You know, addressing education gaps, because the technology themselves are so exponential, there's no way that a teacher can can comprehend that knowledge and share it back to the students, right? That they went to school like 1015 years ago that there's no way that they have that much knowledge in these spaces. So you have to literally be real time education, and plugging that back in. And that's what we're trying to do is provide that mentorship, the ability for the students to get real world experience work on real projects, mentor and be with people.

Justin Grammens :

That's awesome. That's awesome. Yeah, I mean, the next generation is always really fascinating to me, and only with regards to not only the technology, but also the people that are going to be working in it. Right. And so the more you can bring them up to speed and into the fold sooner, the better. It just improves our entire society. So I commend you for the work that you're doing there. And and it sounds very interesting. I guess I'm curious, you know, you've got a lot of different things going on. what's what's a typical day in the life for you?

Jon Herke :

I so thoroughly enjoy What I do, I don't realize where work begins and work ends. I just love learning. I love sharing, I love teaching, I love mentoring. And it all just sort of overlaps. So you know, I wake up, have coffee and do that, you know, maybe do some workout check the emails kind of stuff. I would say most of it is really working with community members, working with students sharing my knowledge, sharing what I know, and getting people excited. I think there's a disconnect. And it's not that people are not smart. It's just they're not engaged. They're not, you know, challenged. I was never I don't think challenged enough in school. To be honest, I was so bored of the regular curriculum. I just we need to get out of the test. Get people passionate, like get them a reason to be passionate. Get them like understanding why they're doing this algorithm why they're learning cosine similarity, when you could be like, hey, let's build our own social network. Oh, by the way, we're going to implement cosine similarity that introduces them how to like find Similar people maybe that these people go to this group and you want to go to that group. So we're doing a cosine similarity determine what group you want to go to. So it's giving some context to that specific learning.

Justin Grammens :

Absolutely, absolutely. Yeah. It really drives the reason as to why you're doing it not so much how you do it. Yeah. And the technology will come in, but there's more, more more people know. And so let's let's dive in a little bit around AI. I mean, the whole sort of purpose in some ways, or a lot of our discussions here is really applied AI, which you've definitely touched on here a number of different ways as you've been out in the community as you've been working day to day with Tiger graph and stuff like that. How do you define AI? I guess maybe we'll start with a little definition. How would you define it for somebody? It's such a broad topic, it touches everything, but I'm always curious to find out. If you were in an elevator, somebody didn't think about it. What's the elevator pitch on on like, what, what's it all about?

Jon Herke :

Yeah, artificial intelligence. I would say artificial intelligence is any thing that a computer system can articulate and provide you it's derived insights. So it is trying to come up with something that it thinks whether it's by you feeding it data, or it trying to learn that data or whatever it is. It's it's, it's trying to is trying to articulate its own thought, or it's trying to derive an output. Right. So I think that's as broad as you can get it. Because there's so many things under AI like NLP computer vision. So it's like, yeah, yeah, I would say anything that sort of like, trying to articulate and come up with an output is really what it is. Yeah,

Justin Grammens :

yeah, for sure. I totally get it. I totally get it. That's big picture. That's that sort of this overarching thing which can be applied to a number of different areas. curious to see as you're as you're working today, day to day you're in there trying to solve problems. Do you have you seen anything interesting that you can share? I guess I know some of the clients may miss if you guys work with or whatever proprietary in some ways, but like how you had are seeing AI being applied. You know, it could Be graph related. Or it couldn't be it could be like, Hey, I just read about this interesting article or this new way that artificial intelligence is being used. And I found it interesting.

Jon Herke :

Yeah, there's a number of paths we can go let's let's start with the easiest one for graph and then I'll go to some other ones that I find fascinating for graph. What's interesting from a perspective of a data scientist, or I don't know, if there's a technical term for a data scientist, what I what I've been finding with those that use graph, they just sort of shocked that they can derive this information in such a fast manner. And what I mean by that is typically the workflow of data science engineer is that they're trying to derive information so they're working on creating all these big joins of big data sets and then they like run them overnight and then like Oh shoot, I got something wrong I got run that again, and then you know, the driving all these different features that they're gonna eventually feed into their model right? And then they get on to graph for their first time and for graph for those aren't familiar. It's a Basically no giants, everything's sort of pre joined. And so you're given sort of a map of the data, then the map is telling you where all the data resides. And then what happens is that when you were writing your query or just accessing places where your data resides, so essentially, you don't have to do the joins, you're just accessing the data. So think of like Google Maps and help your data. And so yeah, they'll run their feature set extraction on on the graph database, and they're just like, wow, you know, I could just drive the feature like, I didn't have to create a derived feature set, I could like literally just access the data and just get back that data for that specific thing. It's all you know, pre joint, so I didn't sit there and wait forever, you know, for it to go through and every single every single row and join things up, I just access the data that I needed. So once the data is accessed, which is phenomenal, and I talked a little bit with Dan, we touched a lot on graph and sort of you know the speed of it and of typical use cases for it. You know, it's not, can't be used everywhere all the time. It's not like a silver bullet. You always want to put it in draft but I guess I'm curious to know once people see this, and they're like, Okay, this, this is awesome. How do you then train models back? I mean, how do you bring in sort of learning into the graph database? Or is that something kind of different? Yeah. So you can take your derived features and then put them as attributes in your graph as well. Okay, so if this is AI, everything NLP would technically be an AI. Dude. Yeah. So I mean, I mean, it's falling under AI. So for example, let's say we just had this one of the students from futures Academy, they took sighs spacey and NLP model that's trained on Allen AI, to biomedical literature. And they took the 60,000 abstracts for the COVID-19. And what they did was they ran it through the site spacey model to do entity extraction. And after they did the entity extraction, they took the extracted metadata and then semantically linked it in the graph, then that allows you to build sort of a knowledge graph. And so then, on top of it, you can put basically like Google search, now you can say, give me All the papers with cancer, right? Bam, because it's semantically linked, and then you have entities and then you have classes, you can go from the class and then go down, you can go narrow, or the other thing is take all the papers, run a cosine similarity or edge card algorithm, and then derive the most similar papers to the one that I wanted to, let's say I found a really interesting cancer paper. And I wanted to be able to find other cancer papers that are most similar to this one, bam, there you go. So there's a lot of things that you could do to like enrich the data that itself using AI, and then use that to either enhance the graph itself, creating more semantically linked connections between the data elements. And then you can also do worry to feature extraction, and then you run it through your model and come back and it comes back with maybe like a pre weighted score that can be used for another thing. So you can you can do you can do it that way. I'm sure the other thing was GCS. I didn't know if we want to talk about that.

Justin Grammens :

Yeah, please. Yeah, if you want to talk about GCSE I was. Maybe you should identify or qualify what That means to people that maybe haven't heard of them.

Jon Herke :

Yeah, graph convoluted neural nets, assuming everyone knows what typical neural nets are, this is using the graph structure as a feature itself. So typically you're deriving a feature set. And then using that to create that model, in this case, you would take the structure of your data and the relationship between your data to then derive a feature in itself. So the structure of your data becomes the feature that you use to then train the model. When you use that you can you can actually strip away all of your features. So let's say we do well actually, I'll give a plug to Parker Erickson. And I

Justin Grammens :

was gonna ask ya, you brought that up that he released something?

Jon Herke :

Yeah, he just created a blog. So it's taking crunchbase data, and he's trying to predict the IPOs. So he's taking the structure of the current IPOs and then trying to find companies. He knows that they're going to IPO but he's going to see if he can predict that they're going to IPO. So the cool thing about the GCN is that even with the features extracted and like stripped, let's say you have no features Except the structure of the data, you're still able to get this really phenomenal score. And so he was, you know, predicting high 80s maybe I think even low 90s percentage of maybe not in the IPL one, but and definitely other use cases that he's done. With GCS he was able to predict a very high accuracy with with even no actual features except that the structure of the data. So I think that's really interesting. That's something that's sort of a new thing. There's been more papers since 2017 2018, I think is when they started really coming out with those GCM papers. I'm sure we can put a plug in the description for some GCS and Parker's blog. Absolutely. Yeah. Yeah. In fact, I'm looking forward to reaching out to Parker here and seeing if he wants to be on a future podcast as well. So how does that differ differ than from something like TensorFlow, you know, would you have any any thoughts on that? I mean, the graph portion does does this get actually loaded into Tiger graph? Is it or is this is something that sort of sits external to it and based on you know, the vertices and you know, The nodes or whatever added are part of the graph. That's what you're sort of looking at, essentially, running, like those kind of algorithms is sort of a graph II thing. You can you can do it in the graph, right? You can actually rebalance all your, your connections, nodes back there working on, you know, actually simulating, I don't think it's actually a better solution. It's so new, it hasn't, we haven't really figured out how to optimize it and like execute and graph and specifically Tiger graph. So what we are doing is it's taking the structure and executing it and a graphics and then you know, you should probably x Parker cuz he's like, you know, really pick my brain to remember exactly what he did on everything. Sure. But yeah, yeah, I would say I would say optimize, like, yes, you can. You can, you can definitely, like replicate it in the graph. Is it the best solution? No. Could it be the best solution and you know, the next year, maybe, I don't know. It could it could be. It could be it's totally feasible. It's just at this particular point is It's so new that we haven't deeply explored it. So it's best probably to do that outside of the actual graph.

Justin Grammens :

Obviously, you've been working in graphs for years. Is this a case? Where? Yeah, I guess maybe it's just the maturity, as these databases are maturing, we're starting to realize, and probably with any technology, right, it's, it's like, oh, we maybe have a narrow fit for it. And once it starts getting implemented and used over time, we're like, holy cow. It can be used for all these other things. And we can start adding in, you know, new pieces on top of it. Is this maybe the next evolution for graph? Or I guess even where do you see graph going in the future? Maybe that's a better question.

Jon Herke :

Yeah. What's interesting as graphs everywhere already, and people don't understand it, everything you do is probably touching a graph at some point. So back in 2012, and Dan probably talked about it, Google came out with their paper on knowledge graphs. And so if you're using Google search, you're using graph. If you're using Amazon using graph using Facebook, you're using graph using LinkedIn using graph. Pretty much everything you touch has graph in it at some point. So where do I see it going? I definitely see. The machine learning space going to be a hot topic just because there's so many things coming out with that structured data and using that as a feature to derive new insights, the integrations tightly deeply integrated with with the system. I see that definitely in the foreseeable future. generalized artificial intelligence is something I think people are starting to talk about. And so if you think about graph graph is a replication of sort of your brain the way that your synopsis are? So I think definitely in the future, it could be a backbone system to a to an augmented brain. I think that's definitely possible. I hope it's on Tiger graph for sure. I'm really into that Jarvis stuff, anything like augmented reality, I fall Elon Musk in his knurling project. And you know, according to him, we'll be able to fully access our brains in 10 years and you know, manipulate our access to data and he's got an interesting perspective on our ability to to disseminate and to interact with data stream. So even us conversing here is a really lossy system. So this isn't my articulated thought I read this in a really well thought out blog article, if I can find it. I'll repost it here because it's a really long read, but it's really nice. And to why he is looking at doing the knurling project to compete with AI because he thinks that AI is going to take over the world. And we need to compete with it, right? We're there. Otherwise, we're going to be like a tree says no AI wants to go talk to a tree just as a human might not want talk to a tree, right? Not really interesting. So neural link, his thought process on this is imagine you and I are friends. Of course we are imagining I go to imagine I go to a movie theater, and I'm at this movie theater, and all these images are coming up and I'm just like, oh my god, this is so awesome. I love this. And then it goes really like scary and dark and you know all these and then it goes you know, happy again and then all of these things I'm feeling and I'm absorbing and then I go talk to a typical Talk to you, Justin. I say just, this is the best scariest movie I've ever seen. Well, you know what is going on in my head is I'm ingesting all this information, huge bandwidth of data ingestion. And then I'm compressing this data in my mind. And then my compression algorithm has to go through a mechanism called speech. And so then I have to like compress it into a word. And then I'm compressing my thoughts into a word. Yeah. And then you take that word and goes into your ears and then comes into your brain, and then you're on compressing what scary means. And your your definition of scary could be totally different than my definition of scary. And so there's a huge loss in data from human to human. And so his whole thing is how can we have that high speed bandwidth between mind to mind and mind computer and right now we're sort of era in a world of augmentation where we're using our phones to augment our intelligence, but we're interacting at a rate of two thumbs. In my case, one thumb. For those of you that don't know, I did lose mobility in life one of my arms. And so that's one of the main reasons I'm a technology enthusiast and looking at brain computer interfaces, understanding how technologies work, some really cool research being done and University of Minnesota where they actually taking brain signals and then converting that into, they're actually having machine learning on the brain signals. And then what they're doing is converting that into an output of the hand. And so you can actually move the hand with your mind so they actually implanted chip in your in your wrist, but they have to retrain the models every so often because they're your minds always reconnecting different circuits. So you train the model, right?

Justin Grammens :

That's fascinating. That's going on here with you. Uh huh.

Jon Herke :

Yeah, we at least I went to a talk there and pretty confident is at the University of Minnesota muscle is just some guest speaker that came in, but it sounded like it was from there, sure,

Justin Grammens :

that whole story about a person have an experience and then sort of pushing data over, you know, through through voice to another person and they're sort of interpretation of it. It reminds me I'm reading a book called 21 lessons for the 21st century, and it's Really interesting. And I think we'll I'll bring it back and talk a little bit about what the author was sort of speaking about. But one of the things that he talks about in that book is while you're talking about becoming a tree, so he talks a lot about just becoming irrelevant, right? And uses the the story of a horse, right. So way back in medieval times, horses were great for travel. At some point, they were actually good for farming. Now we're at a point where horses don't do anything, right. And so the whole theory about this is is is pretty soon humans will become irrelevant, right? That in the future time, everything is going to be automated, all gonna be controlled by AI. And why are we Why do we even exist at sort of any point. But the other thing that and we'll circle back and, you know, ask you a question specifically on that, because I want to get your thoughts on that. But he also talks a lot about just like we understand hormones, we understand human physiology, and what are the hormones that are secreted that make people feel certain ways? You know, yours is different than mine, obviously. But if we know that, based on a certain song being played, I will have a certain feeling a certain emotion and so So how could we then tailor songs to make us feel happy to make us feel sad put us in a different state of mind, you know, to be more thoughtful to be more productive, you know, what have you and an AI could continually sort of train that over time, and kind of put a person in a certain mood that they may have fallen into. But the the science behind it, and the artificial intelligence could make us be that way. So what I was thinking about was, was yes, you, you had this experience at the movie theater and your body chemistry changed, like, how could you use and again, there has to be some sort of mechanism, some sort of mode of transportation to get it over to the other human. But it's like, yeah, is there is there some sort of way that we can use a different and maybe that's what Elon Musk is talking about? is like, you know, right now you're right. It is it is very lossy. But is it literally just like a electrical current that would go from human to human, for example, And oh, by the way, we got the internet, right. So I could do this to people anywhere around the world. Yeah. And all of a sudden, that person can actually really experienced that exact same feeling that you had at the movie theater? Mainly knowing obviously, what what chemical composition changed on your side, and then putting it into what would make the same chemical composition change on my side? Yeah. Does that make sense at all? And is that sort of in the same realm of what he was talking about it? Yeah,

Jon Herke :

I think I think that's totally on the same thing. And what you're able to do, from my understanding with the neuro link project is fire off those same signals. So the same feelings, but the thing is that he said, it's like your hand your hand won't move, unless you will it to move. So he expresses that it's also going to be the same for your augmented intelligence. But the scary part is, you know, I, as we fuse, I mean, everything's probably scary. I'm sure the internet was scary and light was scary and everything that precedes It was scary. The train was scary riding an elevator. That's automatic was scary, because you had the manual guy, but now if, you know I wouldn't trust a guy that was doing that elevator with the little

Unknown Speaker :

maybe he's having a bad day.

Jon Herke :

So, you know, so and then we'll be looking at cars. Oh, you drive a car What? You know, it's the same the same looking at the guy with the elevator like, I wouldn't trust this guy with the car, right. So I, that does scare me though with the having direct link to the internet with your brain. But I do feel you know, you're going to be significantly handicapped with the exponential abilities that you would gain from such a thing. Yeah,

Justin Grammens :

yeah, that's, that's one of the things that happens with a lot of these new technologies is because the person next door is using it. Yeah, in order to keep up. I kind of have to use it.

Jon Herke :

Yeah. It's like saying you don't want to use the internet, right? Because it's, you know, scary, but you know, and of course, you know, you lose your privacy and things like that, because you're on the internet, but there's so much gain from being on the internet. So what do you do? Do you use the internet? Or do you not use the internet? I'm curious to see your thoughts as through the founder of futurist Academy, but you know, what happens in the future of work? So again, we have kids that are learning this stuff, and, you know, as AI becomes more prevalent. I talked about the horse basically becoming irrelevant that you know what, what do you think happens? And what do our careers look like? I guess for people that are, you know, 510 15 years out, I haven't read it deeply into all of the predicted outcomes. But I do believe that we might get to a state where it's purely ran on just enjoying life because we're humans and we want to have technology serve us. So we become lazier and lazier and can do things that we enjoy. I think that that trend is going to stay not to say like all humans are lazy, but they can focus on things that they're really interested in and not worry about, like the minute things mowing the grass and unless they want to enjoy mowing the grass and they come on, mow the grass, but anyway, you have a long day and then you're coming back and you're like, ah, did I really want to mow that grass. So I think technology is going to lead us to the point where we can focus on things that we want to focus on more and have a better quality of life. We do need to fix some things with how we are working With technology and living in this world, I think there needs definitely needs to address the global warming, probably, I guess, sticky topic, but I think there's a lot of things that we could, you know, do for that there's a lot of things that now because we're in this world of COVID My hope is that we understand what's really important we focus on that we start to apply our minds to solving real world problems so we can get to a point where we can thoroughly enjoy life and everybody has that same quality of life that they can thoroughly enjoy. So when we all have the ability to have robotic hospitals or you know, something where something like do total surgery for you, I'm not a surgeon or, you know, expert in that particular healthcare space, but I my understanding is there's less and less people being doctors and surgeons and things like that. So if we need to address that gap, technology's there, how do we how do we get the robots to be able to do surgeries and I already know they are doing surgeries but how do we have like a fully automated surgery room and things like that. Definitely. Going to have some Kind of applied AI and technology that's gonna be in these kind of spaces.

Justin Grammens :

Do you think an AI could do your job today, connecting with people

Jon Herke :

and teaching people, I think that's a very humanistic thing that you need. Because my my job really is understanding people understanding what drives them, what makes them excited, and my journey so far, and I don't know where it's gonna go. It's just it is a journey. And I try to live in the now and focus on providing this value to the world. And yeah, I think I think my particular job is really around that human connection. So I don't know if a robot could really simulate the human connection and understanding humans, maybe they are getting better at that. At least mimicking humans and the human human interaction but truly understanding humans. Yeah, I guess you could, you could, you could take data and try to understand humans but they did that with Twitter and made a Twitter bot that tried them and then went out rampage.

Justin Grammens :

It did. Yeah, you're right. You're right. It completely went went off the rails. So yeah, there's still still ways off on some of that.

Jon Herke :

Not to say that maybe when the generalized intelligence comes here that they'll they'll be like a really a caring robot that's human. Like, you know,

Justin Grammens :

you know, what's interesting is some of the creative fields have always been like, Oh, yeah, I could never touch that. I could never write it write a symphony. Well, it did write, you know, ai could never paint a piece of artwork. Well, it did. And then the thing that I've kind of been holding on to, at least until recently, is, you know, I could never write a software algorithm or write a program, basically, right. We get humans on keyboards typing, like no. So I think I think the whole future of software engineering since it's where my background is, yeah, is is definitely going to change and it's not going to be fingers on keyboards typing stuff out. It's really going to be more on the creative side. I mean, as I work and more and more like TensorFlow and stuff like that. It's really dialing the knobs on the outside to try and figure out the best thing and letting the computer Do what it's best at. and optimizing algorithms. You know, I think about just like SQL database, you know, writes Pretty soon, you're just gonna speak find me this stuff and it'll go out and

Jon Herke :

now you're just gonna pick it up. It's like, it's like, okay, there's a calculator there. Do I want to do it by hand? Or do I just want to put punch some numbers in on the on the calculator? Oh, no, that reminds me of doing work. So one of the students was saying that Oh, man, I got to do this history paper. I was like, Well, why don't you use deep fake to, you know? Literally, you can you can have the deep fake write your paper and submit it and see if that actually not and, like you can't get mad at a student if he articulated and executed a deep fake paper and it was so realistic that they actually got accepted. And I mean, I would give the kid that some credit for that, you know, right.

Justin Grammens :

Actually digging in figuring out how to stand it up and get it get it working. Yeah. Or even thinking outside the box. Yeah. How can I optimize this As best as possible, have a computer do it.

Jon Herke :

Yeah. Yeah, yeah. So it's it's definitely interesting.

Justin Grammens :

So yeah, I want to give you a little bit of time to plug your stuff here at the end. I was just like, just a couple more questions. Are there any classes that you advise people taking as a as you've done over time? Any books, you know, conferences, stuff that you find interesting, I guess. And then I guess I'll let you talk about how people can reach out and connect with you as well. But yeah, I'm just kind of curious if there's things that are Top of Mind someone's getting into this field, you know, graph or non graph, you know, emerging technologies? Or what are some things that sort of come to mind that you think are interesting?

Jon Herke :

Yeah, I would definitely try to figure out what are the major influencers, try to find the people that interest you that are in the space that you'd like to follow. Also, creating a network of brilliant AI mentors is really important. So I'm sure anybody from this group would definitely help anybody that's interested in AI, no questions asked. I would say 99% of humans. If you ask for help, they'll say yes, there's very few humans that I know Actually, I don't know any

Justin Grammens :

so great. 100%

Jon Herke :

I mean, I I try to keep people positive. Yeah, so that's a pretty good pretty good chance if you ask somebody literally for help and and it's not like you want them to solve your thing, but you're just generally curious about something or just want some mentorship. Yeah, that they'll help like so I would say definitely surround yourself with mentors, surround yourself with like minded people find people to learn with find people to grow with to be excited, I just built this and share it, you know, and they get excited. And then like, share it in their thing. I would, I would say mentorship, find those communities and then just explore things that are interesting to you. I do find value in curriculum like going deep in courses, but I do find a lot more value in just sort of exploring exploring that first like tier of all these different things. And then when you get to a point where you're like, oh, shoot, I can't understand it, and then you're trying to go deeper a little bit, then it's really important that you go and find that the right course to go deeper. I would advise not to learn about the whole ocean You'd be riding your boat on the top of the ocean. And then once you find like an interesting shipwreck or whatever, you just dive deep, right? So yeah, you don't need to learn the ocean just to sort of explore, play, you learn. And then once you see you, you find something you're passionate about. That's really intriguing, really interesting to you try to go deeper find that find the book, the curriculum, find the mentors. That's great advice. Great advice, john. Yeah, totally. Totally. So how can how can people find you obviously your LinkedIn, Twitter, you know, what have you and I'll include all this stuff in the in the liner notes, too, as well. But yeah, what's the best way for people to reach out to you, I would say LinkedIn, LinkedIn as probably the number one place that I am in 24, seven, and so if they want to connect with me, just look at the link on this. I'm sure the link will be there. Otherwise, search my name john Herky Jo n, h er ke and then send me a request, shoot your name and then say a little intro to who you are and except your else except your Connect. So and then we can talk and we can relate. Stop ideas.

Justin Grammens :

Great, great. Awesome. Well, thank you, john. I appreciate all the time. It's been a great conversation and wish you all the best and I know you'll continue to press us and, and do some amazing things here in the community not only around graph but you know, helping helping kids get more into this with the futurist Academy and everything going forward. So yeah, thanks again. I appreciate it. Thank you, Justin. All right, take care.

AI Announcer :

You've listened to another episode of the conversations on applied AI podcast. We hope you're eager to learn more about applying artificial intelligence and deep learning within your organization. You can visit us at applied ai.mn to keep up to date on our events and connect with our amazing community. Please don't hesitate to reach out to Justin at applied AI dot m n if you are interested in participating in a future episode. Thank you for listening

People on this episode

Head Shot

Justin Grammens

Host