In This Episode
Artificial intelligence and machine learning are two of the hottest and most important concepts in the tech space, but are we actually close to a point where computers can make autonomous, non-algorithmic decisions?
This week on the Veterinary Innovation Podcast, Shawn and Ivan are joined by Dr. Aubrey Kumm, the founder of GuavaVet, for a discussion about the differences between artificial and augmented intelligence and what you need to look for when hiring developers.
Dr. Kumm recommends The Lean Startup: How Today’s Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses by Eric Ries and Hot Seat: The Startup CEO Guidebook by Dan Shapiro.
- Transitioning Away from Being a Veterinarian
- How to Find Great Developers
- The Difference Between Artificial Intelligence and Augmented Intelligence
For everyone here, from Veterinary Innovation Podcast, we wish you and yours very happy holidays.
Ivan: Hi, my name is Ivan Zak, and I’m happy to introduce Aubrey Kumm who we got introduction actually during the podcast from Brus. So today we gonna talk about artificial intelligence in veterinary medicine. Aubrey is the CEO of Guava AI. He’s the founder of Guava Vet, which is a recruiting technology and he holds a degree of master of medical sciences and developmental neuroscience from the university from Cape Town. He also holds a bachelor of veterinary science in veterinary and medicine from the University of Pretoria. Was one of the first vets in Europe to be credited by the European school of veterinary postgraduate studies with the general practitioner certificate in endoscopy and the surgery. And recently, had a terrible airport experience in Manchester. I wanna hear about that before we jump into all the cool stuff. Aubrey, what happened?
Aubrey: Well, thanks for having me. I was actually in Manchester for the private innovation summit which is a veterinarian innovation summit organized by the Royal College of Surgeons in the UK. And, well, I was only there for a day, because at that point, I needed to be somewhere else early in the morning. But when I arrived at Manchester airport, there were quite a number of ques and quite a number of people waiting for the bags at security. Now, the strange thing was, is that usually, you’ll see one or two bags will be sequestered for further exanimation. But it was if only a few bags went through. And so I sat there, and I waited, and I waited, and I waited, and you know, getting to a half an hour. And I just saw that more and more people just waiting for their bags to have secondary examination. And at some point, I just walked to the one guy who was, you know, doing his work, and you don’t want to interfere, because people’s insecurities are very important. And he, I just said surely something must be wrong, cause everywhere else there is no this problem. And at this airport, every other bag is actually being sequestered for a secondary examination. Isn’t there something wrong with the system? I was obviously told to, to sit down and wait, and the bad experience is that I had to run has been to the airport two hours before, I had to run to catch my plane. So that was it. It’s a, and, you know, I think I’ll say something?
(Ivan laughs, indistinct speech)
Aubrey: ..experience, that one happened.
Ivan: Alright. Well, you did catch your flight.
Aubrey: Yes, in the end, I did catch my flight, one of the last people on board, having all these people rabonaking as you walk past. But, you know, that’s one of those.
Ivan: Well, that’s quite a mixed experience. And I really would like to know more about how do you, how you get from veterinary sciences to neuroscience, and then how do you mix the two in your current ventures?
Transitioning Away from Being a Veterinarian
Aubrey: So, I had my own practice for about good on 12 years in the UK. So I’m born in South Africa, went to school in South Africa, I then went to the UK in 2007, I thought I’ll stay there for a year, I ended up staying for the 13. And I thought of, this was great, and I did everything I wanted to, now it’s time for something else. And I’ve always been very interested in the brain. And how the mind specifically works. And that’s where the interest came in neuroscience. And, it took about a year to get to, involved in a program that I was really interested in. And that’s how I ended up at the University of Cape Town. Specifically, in the department of child and adolescent psychiatry. And my master’s and thesis was on and called “Autism embryo”, which we developed in collaboration with Duke University. And basically, what this app was for, was to enable people at home to screen their children for autism. And, to be able to do so, no matter where they are, using their mobile phone. So we tested it on the Apple, iPhone, because, you know, Apple was part of this project. But that’s how I got into the, so I’ve always liked technology by doing, I did the ophthalmology course as well with the postgraduate studies in Europe, and also the endoscopy and surgery. And there is a thing there. And a thing is to see things better, what you need, especially when you get to middle age, is illumination and magnification. Because if we can find an easy and more effective way to do things, we will save time, and if we save our own time, it means that we have more time with our plans, and our patients to do something we can do to less the human conflict. And if you consider the autism on the content of Africa, for example, they are half of billion children or young people under the age of 18, one of the hundred things say or between one and sixty or one in a hundred will have autism spectrum, and they are only fifty child and autism psychiatrists, and or psychologists, on the continent of Africa. No matter how we try, we will never ever be able to produce enough specialist people to deal and to help all these families who deal with autism on this 05:38. So what do you do? You need to find a way to scale up and scale-out. And this is where the app came. So what we can do is we can have these people in the first instance that scream for the?05:48, and then using the funds as well, provide them with the next steps. Because ethics, it would be completely wrong, I think, to say “Look, there is a risk in your child to get autism, but that’s all we can do”. No, what we need to do, is no matter where you are, say that these are the next steps you can take. So not only do you scale-out, in other words, more people in one area that you scale up as well, in other words, more people in different areas and even different professions. So that is how I got into from veterinary medicine to neuroscience to the development of artificial intelligence, and how that can help us with basically providing services to people.
Ivan: Wow, so that’s fascinating. So you know, this is really close to my experience in, you know, a transition from vet medicine to developing technology, and everybody has their path into it. And you know a lot of people come out with the great ideas, and then they, I’m sure, well, I don’t know what your sophistication, but I don’t know if you’ve heard it from your colleagues, I heard a lot from my previous projects, SmartFlow, I thought about that and I wanted to do it. And then, I don’t know if you’ve ever heard anything like that, but what does it take people to move from “I had this great idea” to actually move forward and step into unknown to change the status quo or to develop a technology when your background is in the veterinary medicine, neuroscience, and then now you’re developing a technology. So how this transition happens, how do you actually move toward that next step and say “Okay, now we need the technology so let’s incorporate it into my master’s and everything I learned there”. What was that journey for you?
How to Find Great Developers
Aubrey: Firstly, I know what I can do and what I couldn’t. And I know where I will need or where I will need help with other things. So, I don’t code, that’s not what I do. Nor do I want to learn to code now. But what I like is that I come up with the idea. When I have an idea, I have a group of people. Friends, and family, and people with more or less same interests, and then I sort of, we start talking about this idea. We govern it, for example. So I was in the flight back from, this was, I think it was San Francisco. It was in the US, and I was at the conference for, for, it was an autism conference. And on the way back, I had this idea of, if I didn’t apply my own app, and it would be to get to connect people. And you know I didn’t come up with the idea on the flight, but then back home to South Africa with friends, and fam, and it was, you know, I was jetlag, and I didn’t sleep all night, and I woke up. Well, I was still awake at five o’clock in the morning, when I heard the other people are awake. It is like a turn of event. Firstly, I went to 08:31 and said “Look, it’s matching. It’s matching people with similar interests to form a relationship. So, the questions. No, no no: there are people who want jobs. But they want jobs in a specific place and specific practice. And then there are people, the practice owners, who emphasize. And if you want, if you go and look for jobs at the moment out there, it’s like to try to run from a firehouse. The information is all over the place and you have to go and look at a lot, you know, of different situations. And this is my idea and you know, so I ran the idea first thing, and, you know, we started talking and talking. He liked the idea. Then, I started researching and looking at whether or not there was anything else. So this is a very similar story, I think I’ve heard it in your previous podcasts. This is what you do: you search if there is anything similar out there somewhere. If it is, good, it doesn’t mean you need to stop. It means “can I improve this? Is this something I can do better?” And then, once you have a better idea, then you start talking to the developers, and this is your next step. And this is a very, very important thing is to find the right developer. Now, if a developer that just says yes, is probably not the best developer. My developer challenged me all the time, and this is what so excited. We’ve got this, we’ve got this thing we do whenever we add new attractions to the new platform, are we give it stars. So we started, if I go back to recruitments, what does one store recruitment platform looks like? Well, it would be the adds you would find in the newspaper. And then we worked up to about eleven, twelve, thirteen, fifteen stars, even now we see 09:58-10:00. See, the future, we probably have around seven of eight of them. So that’s what you do: you discuss with people, you find the correct people that can do the things that you. I’m surrounded by people that are much more shiver than I am when it comes to these things. I’ve got this job, I think of the ideas.
Ivan: Wow. You know, this pretty much how I look at the development. Cause none of the things that I’ve built is something that I could’ve built. And the last company and developers like to remind me of that, cause in the brainstorm sessions, I remember I said to someone, I said “You remember when we built this feature?” and one developer turned and he said to me “Excuse me, you didn’t develop any features.” I said “You’re actually right. I thought of them, you guys developed it, thank you”.
Shawn: It’s awesome
Ivan: So it’s a, so I remembered that careful, and now in our meetings, I say “Whenever you guys made this”, that’s an important correction there. That’s fascinating. So one I think, I wanna kind of dive a little deeper, and I don’t know what’s your experience looks like. It all sounds so good when other entrepreneurs say “Oh I had this idea, and I found these people who are specialists in this” But then how does actually this journey look like? How does one physically go? We talked to Stacy Santi, she built an app, and then she, you know, she went out, and how did you find them? Where did you have the connections? Did you have (indistinct) coded? Did you have funds? Did you have to raise capital? What was this story like?
Aubrey: Oh my goddess. Look, I think if I tell the story, nobody would want to do what we do. Firstly, no. At this moment in time, funding is friends, fools, and family as they would say. So most of it is myself and one of my really good friends and colleagues who’s also a veterinarian, specialist veterinarian with regards to the production. And we basically, I first started, and I said “Look, do you want to be part of this?” And yes, this is the funds we need. Let’s start with this. But initially, it was itself funded. Then, you get the idea. And you start developing whatever this technology is that you develop, and mine is an app. You know, at some point, it was a disaster, cause the initial developers that I used, were keen on me, just adding and adding features where the guys that I’m with now. So much more professional. And, you know, it’s a team, and they really guide me. And that’s not easy to get to. It was my chance that I’ve got these guys. And, I mean, I did what everybody else would do, just talk to Google. And I started with Google, and I looked at ratings, and I looked at the area I wanted to be, I looked at costs because too cheap is not good, but too expensive I simply can’t afford. So it was a very long process to get to these guys, and then also during development, we suddenly realized actually what happened. Apple changes its rules. And I was ready to launch at Vet show in 2017-2018. And a month prior to launch, these guys 12:39. I thought something was wrong, because, as for staff to be ready for the London Vet show, and it just didn’t happen. And then I got this email, not even just a call. Email saying “Look, we’re so sorry that your dream is shut and so is ours because Apple changed their rules, bla-bla-bla, your app 12:50?”But I don’t understand. I specifically asked for you to develop a native app. In other words, from scratch, you write code. And that’s what we started with “Where are we now?”. So thanks goddess the guys that I use now was sort of you know part of the returned project, and I said and I could continue with them. And that’s why I have these brilliant relationships with them. It was very easy, and I think I probably aged there in ten hours when I got that news. Because you ready to launch, you know, it’s two months or a month, and these guys say “Look, your app is not gonna happen”
You know, what happened with the previous code? Where is it? It couldn’t just disappear”. So no, that is definitely not an easy step. And then the next step funding. You get to a point where the money, how do you bankrupt yourself? Well, it’s sorely and then suddenly. You know, that’s how it works. You need to be careful all the time to make sure that this is funds because it’s 13:41-13:42. There is always something new, or something else, something better that you need to do. And other rules you can read all the books, which I have. The “Lean startup”, and “Hot seat”, and “CEO”, and you know, all these books. And they give you an idea. But most, most of these books I read after I made these mistakes and so yes, 14:00-14:05 So you know, I think the first thing to do is if you have the idea, start dreading soon. Do your research. You know, speak to clients of these developers who’ve had technology developed, you know. If it’s a native app you want, then obviously, depends on, that is a different time of developer who simply will use a plugin play model for you, and they already developed something which you can just adapt for your own use. It’s a long and hot road, and it’s like anything. You know, when I started my careers of it, and I worked in locum for about a year, I never had anybody working with me, because it was in a day of the foot and mouse in the United Kingdom, and then I was qualified, and I had my own practice with it. A friend of mine, who bought an app, and we had to learn a hard way. I never worked for anybody as a boss, and I had to learn how to be one. That’s one of those things, you jump and you learn to fly on the way down.
Shawn: That’s so, so funny. I totally know what you mean. I wanna direct us a little bit toward this artificial intelligence area of your business. These episodes are so short. And I wanna get into meat and potatoes into how you’ve leveraged AI because as a technologist my entire life, I love the way that I’ve heard of AI, kind of referred to, and me and Ivan were a couple of guys from Google couple months back, and he worked for one of the largest AI companies in the world, and you know, our AI conversations centered around you know, spreadsheets and PowerPoint. And I’d like to kind of understand how you leveraging it in the software that you guys are developing.
Aubrey: Firstly, do we talk about artificial or man-made intelligence?
Shawn: Yeah, yeah, so I mean you take us in either direction or both.
Ivan: No, define it for us, please. Because I don’t think we are certain about that.
Shawn: We need more intelligence, that’s the one thing that we are certain of.
The Difference Between Artificial Intelligence and Augmented Intelligence
Aubrey: I would say, at least my understanding, at least, is that artificial intelligence if we take artificial intelligence versus augmented, is that artificial intelligence is almost seen as replacing human intelligence whereas augmented is more of a collaborative process. In other words, I learned a machine to learn, and we both benefited from this, rather than being replaced. So what we are doing now with the platform, is like with any intelligence, you need data. As we collect data, which is what, how you build artificial, how you move augmented to artificial intelligence, it has to do with deep learning, the more data you get, you have an algorithm. You can certainly start developing a technology that helps whoever is that you need to help. So in this case, a practice owner needs to find somebody to come work for him, to do a specific job, and job seeks wants to go work in a specific place, doing a specific type of job, and what we are doing, is we are saying right. If we collect all the data. So, for example, whether people are how far they would like to travel, any job that you have on our platform, which is words, you know, little description, and image, for example, the image of the practice owner. To people at the practice, we say “Take a photograph”, and this photograph needs to in a very obvious way, try and show your practice culture. So what we can do now, we can collect data, say “Okay”. People in this area want to travel only this far, and this is the amount of money that they would like to receive for the job that they are going to do. The average salary in the area is this, and this is the type of job they want to do. They don’t want to work out of hours, they don’t want to work by themselves. They do want extra support. And if they look at the photograph, you know, what type of photograph. What was the hook? You know, if you take a photograph of the building versus the photograph of the dog versus the photograph of the people having fun, you can see the community inside the practice, or the culture inside the practice. So the idea is that as we collect all this data, and we analyze this data, that as somebody completes all the fields to post the job that they will help? That says “Look, if you do this, this is probably how long it’s gonna take”. Or this is what you want, this is what you’ve selected in this area. But if you do this, this will improve the chances of finding this match, finding this person. And then basically, to create this one, one platform where people can meet each other, and make it easier by using all the data out there, from the outside. Connect people, and that would improve, they would start with the better foot. In other words, the relationship between the employer and the person coming starts on a better foot. Does that make sense?
Shawn: Yeah, it makes sense. But to me, that’s just a fancy filter and a spreadsheet. You know, assets of functions of the website, and this is my issue, and I apologize
Aubrey: No, of course
Shawn: But this is how I look at artificial intelligence in the state of machine learning. We’re just creating fancy algorithms or fancy filters, you know? I don’t think we’re actually to the point where we’re actually having intelligence that learns on top of itself that you know, makes advances. And I’d like you to get a comment on that.
Aubrey: Absolutely. That’s where we are now. So now let’s say. So if you develop artificial intelligence, it means that you need more data. So now let’s say for example you use our platform, and our platform could start connecting with the practice management software. So as it starts progressing, and we take data from, you know, a number of inputs, we can start predicting when you need the next person. When you need extra help. And now, look, I’m going a couple of years ahead from where we are now, we just collect data. In the platform, in the app, we have nine specific skills. So we’ve got no specific skills. So working dates and cases, diagnostic imaging, and these, nine separate ones. So we can also start learning, we can imagine, because exactly what you need. If it this time you need more surgery, we could predict that you need more surgery, and there will be a point where we can say “Hi Siri, I need a vet on Monday”, moving to whatever is a machine just knowing that you need a vet on Monday, and a vet may just arrive. So that is where we are. So now I’m talking fifteen stars. Yes, we definitely not. What the algorithm is doing is now we’re gathering data. As we start gathering data, at some point we will be able to have this machine that will help you recruit the best person, and this is very important. That the way we work now is somebody comes and works, and you’ve got contingent for staff that works on a permanent basis. Well, that is not how the economy is gonna work forever. There will come a timer where people will want to work, you know, an hour here, two hours there. And it’s a gig economy. There’s no reason for us to believe that economy will not be a part of the veterinary profession.
I think you’re totally right. And, you know, another area like to kind of interject here is you know, I heard within the last year, for the first time in the history of the world, data has not surpassed oil in value. And you really built an interesting point around artificial intelligence and the state of it today. And the state of it today is that we collect the data to solve tomorrow’s problems. Does that make sense to you?
Aubrey: Yes. I would say that we’re collecting data to solve problems we don’t even know we have yet. And that is actually the truth, yes…