Artificial Intelligence in Business Dan Yarmoluk – Episode 19

Artificial Intelligence in Business

Since there is so much hype around artificial intelligence examples in business and the scary misinformation that was spread around this a few years ago I decided to do some research and interview some experts on artificial intelligence, machine learning and data science, IoT etc…

Dan Yarmoluk- IoT professional and adjunct professor, has been in industry for many years and after many years of study into machine learning and ai (artificial intelligence) is now teaching machine learning as it is a passion of his. He is also delivering artificial intelligence and working to implement it into business units, including manufacturing, supply chain infrastructures and other areas, to solve complex problems and save money over the longer term for businesses.

In this 30 minute podcast Dan  and I cover some real world examples of what Artificial intelligence is being used for now and in fact take the time to dispel any myths that you may have of the robots coming to get you and to take over.

Artificial intelligence examples in business are a bit thin off the ground to be honest as the people who are using AI are on the whole pretty quiet about it.

Nathaniel Schooler
Well, hey, Dan, it’s really great to speak to you again, I’m quite excited, actually, because you and I share this love of automation and AI and data, clearly, you know, a lot more about data than I do. You’ve been studying data science from what I, what I understand.

I just find it so exciting, like the world that we’re living in, and there’s all this doom and gloom, you know, about us losing jobs, and we’re going to die and, you know, and all this kind of stuff.

If we’re keeping an eye on what’s really going on, we should be able to stay ahead, right? And stay upbeat, and actually create really exciting jobs for people, reduce the working hours, because we’re all working too many hours. A lot of people are doing things they don’t enjoy doing. So what do you think about about that?

Daniel Yarmoluk
I think it’s going to solve, you know, for me, you know, one side of the equation is the jobs and, you know, as they is the status quo today, but what’s interesting to me is, what solves problems can we solve tomorrow, I mean, a big one, I’ve had many people in my life affected by cancer.

With these tools of mapping the genome or micro targeting certain kinds of things, and having a better understanding or the idea of, you know, the computer being able to go through a database of millions of tumours instead of a radiologist.

Just think how, how many photos a radiologist or tumour types, how many photos he can kind of absorb 1000 maybe, which I think is probably way out of line and understand the nuances versus a specific doctor nailing a certain thing and then you can leverage that information,

I think, will provide us just much more exact customised solutions for the masses. So I’m excited about about the solutions. I opt for that over a disruption in certain segments of the market, you know, so I’m excited. I’m super excited to be participating in it and being an observer at this time in history.

Nathaniel Schooler
Me too, I am excited also, so you’re, you’re focused on kind of automating tasks and things like this these days, or because you have a background in data science, and now you’ve moved into industry.

Daniel Yarmoluk
So my background was an MBA guy selling kind of data solutions for decades, globally and then I just had this interest in, I kind of audited some classes and data science, and then got my masters.

Now I teach in the graduate programs, around software so I’m a late I’m late comer, but come from a different angle. So I think that that has value and yes, I mean, industrial kind of automation or predictive maintenance, predictive analytics, looking at the sensors of, of rotating equipment.

Artificial Intelligence in Business is so Exciting

We look at all the data that’s coming in all these different kinds of leading indicators and lagging indicators and weak to get is when you put them in that constellation, you can get a predictive element so that we’re getting more productivity out of a particular machine. So it is an industrial play. But it’s a it’s interesting to break down specific processes and try to find automation steps in there. They’re all micro steps and baby steps, but when you add them all together, it starts becoming quite interesting.

Nathaniel Schooler
Right. So So do you think we’ve actually so there’s a lot of hype right around all of this. Yeah, you know, as well as I do, people are like saying, they want AI for this and that and everything else. Well, has anything really changed since I sat in a meeting in 2014, 2015, I was invited to sit with IBM around a round table with the editor of Wired and all these people.

I was quite worried about the disruptions in jobs then and then in 2015, Hitachi hired the robot warehouse manager that increased productivity by by 9% or 10% in 24 hours.

It took over managing humans, Oh, well, let’s just get a bit of advice from a robot. It was like manage humans is how they put it. I don’t think any, any things really changed since then.

There are a few more test cases, you’ve got Wimbledon who pioneered in 2015, the AI using the database to create the real time updates of how fast the shot was. If it was the second fastest shots in the history of Wimbledon. So they based all these things on the databases that they had, right.

It’s all about the database, isn’t it and then taking that database entering it into the AI and then asking the right question isn’t it’s not a magical formula is it?

Daniel Yarmoluk
A lot of the backbone is data and this is what’s kind of interesting about the book by Kai-Fu Lee, AI Superpowers: China, Silicon Valley, and the New World Order. China has a new world order in that you need a massive amount of data.

So we got, we got a couple different things when we talk about artificial intelligence, we have got machine learning and deep learning.

Nathaniel Schooler
Can I just stop you there, what’s the difference between machine learning and deep learning.

Daniel Yarmoluk
Well, deep learning or neural networks, as opposed to emulate, you know, various kind of mimicking the different kinds of neural networks in the brain and how it has certain levels, that the computer can emulate the kind of different filtering mechanisms and interpret things.

In machine learning it can be rules based and a little bit more technical and doing things like classification of spam, or not, never sent an email to this person, it looks suspicious, therefore, it’s spam.

So I think there is some ways that AI and machine learning in terms of classification. Which is a very defined set of variables and parameters and then being able to decipher is it A or B. So a classification mechanism and it can be kind of complicated.

I can give you an example where I have a contact and friend and one of the biggest patent attorneys in the world has created a paralegal process that he is automated. So in intellectual property, there’s all this queuing up of documents called so they call that docketing.

This goes through either the American EU or Asian regulatory bodies where you file a patent and then you have so many days to or months to file it back. These are hundreds, or thousands of very specific documents and thresholds to get your patent application accepted.

All these gates well, pretty sophisticated educated paralegals were doing this for years. But the accuracy becomes difficult because it’s a very mundane task, you got to have a specialized skill at the same time, it’s very boring. So it seems to be a recipe for disaster or you have the most librarian type mind.

When you’re doing that on mass as the number three this person I was referencing, is the number three patent attorney globally. You know, he was a machine and he had to, he looked at, like, how do I if the people aren’t paying for these paralegal kind of functions?

How do I automate this? So first we went through that he went through that period in Indian offshoring and using Indian talents, but the accuracy and Indian labour rates are going up such that, the turnover was high, so we couldn’t train them fast enough, as well as human accuracy issues.

So he automated baby steps, baby steps, baby steps, and then it went, you know, longer and longer and longer to be an entire documenting solution. So all that documenting paralegal work and if there’s exceptions to that, you know, the, all the variables don’t work, it spits it out, and then the paralegal looks at it. So accuracy, speed, and more importantly, the lawyer can work on more strategic issues, should we be filing this patent? How does this is patent look in the constellation of Asian AI patents out there for as an example.

So I always see this as a great thing, but it will change where people are working, there is going to be a disruption in labour. But the question that you and I have is like, where do we focus our efforts? And where do we educate our kids for the future?

Nathaniel Schooler
Yeah, very much so, I mean, I think that the most important thing far as I’m concerned, it’s just making them adaptable. They need to learn how to adapt and pivot very fast, and they need to learn and be continual learners, instead of this, oh, I’ve got my degree, man, I’m going to go and get a job, this ego problem, people have you say, finished university jump into a job, and they think they know everything.

A lot of people find that actually, they don’t know anything at all. That, that, that’s when you begin learning is when you realize that you know, nothing, right? I mean, I forget, the wise man said that. But that’s my, that’s my philosophy is why I launched this podcast, because I get to talk to amazing people, like you.

Daniel Yarmoluk
I try to keep up with the amazing people. I mean, so I think it begs the question what, you know, what are our best human elements? And I think it’s the what questions to ask, right? I mean, the computer can do the heavy lifting and granting of databases. But that doesn’t mean the question that you’re asking is the right question or it’s a meaningful question or problem to solve.

So, you’re right. I think the human has to be adaptable has to be open to different approaches. The thing that’s interesting, it’s not just about AI, right? It’s about all these different technologies accelerating at the same time, you know? And as Peter Diamandis calls exponential technologies, they’re converging.

They often look kind of deceptive. They look Snapchat ish. They look like young demographic and cutesy. But what really is rocking it is the underlying ways in which people communicate how easy it is to send files and paper and solve things and automate certain things. So I kind of wrap it up in digital transformation or cloud computing or quantum computing, but as you see robotics 3d printing, material sciences, all these things are witnessing exponential growth, which which go into this cosmic mix.

Nathaniel Schooler
I just held up my, I’m sorry, I hate to do that. But my friend Erik Yeah, he’s really into his tech he bought a 3D printer. One day he gave me this Wanger a 3d printed knob.

Daniel Yarmoluk
Just for the audience that’s a 3d printed middle finger.

Nathaniel Schooler
is, this is not a knob, I would I put that in my drawer or I might have sent it to someone.

But the future is bright, there are so many opportunities out there to kind of marry your enthusiasm for life and your skills with technology. But in my in my mind. I’m actually sitting back and I’m thinking, well, it’s really about the companies and the people who are leading the companies and what their ethics are. Yeah. And actually, how they’re looking at the world.

They need to say, well, okay, we’re not going to automate this particular thing, because we think it’s going to be better with a human doing it. Because it’s pointless having a business that’s got two employees, that makes $20 billion a year, right. So so then it’s like, well, what, what gifts can these people bring to the world through that business? Then how are we going to use the AI (Artificial Intelligence) to sit alongside them to make it better.

I was talking with David Mathison who I know you the CEO Summit founder, yesterday interviewed him. He was talking about this new book that’s come out, which actually talks about how, rather than the second Machine Age and how you should actually give people universal basic income, which to anyone that doesn’t know that’s just like meeting the needs the basic living expenses of everyone on the planet. He said, rather than doing that this this gentleman, I forget what the name of his book is, actually, I’ve got it written down. I’m going to interview the chap fairly soon. But David said, I need to, like learn a lot more before I sort of speak to him. It’s called I think it’s called team human. Is that what is that what it’s called the book.

I think that’s what it’s human. I read almost every book that I can get my hands on. But

yeah, I think that’s what it’s called. It’s not quite out yet. But so so. So what the guys saying is, rather than actually doing the universal basic income, that we should not give them income, but give them a stake in one of these tech companies, so that then they can actually grow and then make money and own shares in the business and then grow with the business. So it’s a bit like the Blockchain thing of a new social network and giving away shares in the Blockchain social network and saying, well, you bring in X amount of people, and then you’re, you’re gonna you’re good actually grow your income in the future, instead of having something that has no employees, you know.

Daniel Yarmoluk
I’ve also heard statements like, well what is the right task for humans to be working on meaning maybe it’s rethinking our idea of value, that if we’re having an ageing population and more of them are living longer caretakers because it’s a very human thing. Let machines do machine type things, and valuing and compensating people for taking care of the elderly in a different way, which is time love concern. So those are interesting, interesting, you know, thoughts. And I think we’re going to be reckoning with some of these things. But the ones that I feel that are going to win in this new world is, is solving people’s problems.

So, I mean, the market will basically speak and I think the dissension is happening where people were looking towards the future, they sense it at the same time, they have got to, cover their store or their p&l today and it’s a big reckoning between what the new is and what the old is and that’s the disruption of these things.

In the United States here you know Sears was a big department store you see that sign going down is it all e-commerce where they’re where they ignoring the market signs? I mean, there’s probably a whole there’s a huge story there, but things are changing and, you know, millennial buying patterns, patterns are changing what they value, and quite honestly, I don’t think the status quo has been equitable for for everybody, as is. So I think the shaking it up is is potentially a good thing.

Nathaniel Schooler
Yeah, I agree. I think I think I think certainly, in terms of in terms of mediocre service, mediocre brands, mediocre quality, we’ve become a nation certainly don’t know that in America. But we’ve accepted the fact it’s okay to buy, buy something that you feel vaguely happy about. Like, it’s okay. That, you know, you buy something and it and it breaks, you know.

I bought like a tripod, right? For my, for my, for my phone and it broke like, within two days. No, it was actually for my for my camera, it broke within a week. I found one in a charity shop for the same price it’s like 10 times the original price. Yeah, it was like, you know, a 95 Pound tripod.

I’ve still got it, I, you know, I could hit someone over the head with it. And it would still be right in 15 years. I think, actually, that we’re going to become better at looking after the high quality products.

I looked at a video from one of my, one of my friends the other day they’ve been in Grenada and seen a distillery, because I mentioned my Godfather has invested in a new distillery out there, so they are sourcing the cane from the field.

Then they’re actually fermenting that one that one bit of sugarcane that particular day. Then they’re doing full traceability of the product, right, all the way through to, you know, temperature controlled storage, and then, you know, all the way through right to the shelf, it will be I would think

That kind of things quite exciting. But then I looked, I looked at this old piece of machinery, and my dad would have loved it. It was like some kind of water powered crushing machine that pressed all of the all of the sugar cane. And then and then there’s this like little little channel whether whether river of sugar cane water comes through, and then it goes out and it goes into this tank. If we can if we can help to tell those stories of these amazing products and amazing places. I think the future is very exciting.

Daniel Yarmoluk
I think the big boys are winging out because they’re there they’re, you know, these things like supply chain and understanding where our ingredients come from food. If you have that Blockchain traceability, I know they’re starting to do with Walmart, where we can understand that if meat comes from that farm, I mean, that’s what we’re willing to pay extra for.

So I think it’s it’s democratizing, but at the same time, it’s shaking up what we thought was normal, there is no new normal and when we talk about being accepting of change, and being comfortable with ambiguity, I’ve heard is the statement, you know, comfortable with not knowing. I think of it as a special forces team, how they got to take all this inputs and react. I think that’s how we have to start thinking about things.

As a father of a 15 year old daughter, I often wonder how this 19th these post World War Two infrastructure factory of education is servicing them for the future. I just, especially when you’re talking like hardcore math and science, I’m a stem guy, I’m a computer guy, believe me, but you know, I don’t know if the algorithm does that. If it’s justified it like they should be applying it not like learning over and over again, for the test. You know, so I’m nervous about it.

As we debate in the west of what’s going on, I mean, China’s rising Chinese hungry, they have so many billions, they got 1.3 billion people, and they look at, you know, iterating business models. The whole concept of them as a copycat nation also has learned it well, to this time in history, in the sense that they hyper iterate on a Twitter like thing in China, add a payment function, and then all of a sudden, it went from a cashless society to a cashless society. Because everything we can be paid on WeChat.

Reminding you, they have no regulatory issues with the data, because they’re trying to get ahead with governmental support. Wow, all that engineering talent, the government’s investing trillions of dollars, as well as opening it up, and the data is free as we’re debating and wrestling with regulatory issues. But that’s also ok. Right. Like, we we can’t assume that we own the world either, you know, there’s a balance of spatian in, democratize communication. So it’s just going to be a very interesting time, and I think we will solve very complex problems. But, you know, there will be a shake up in the distribution of income, though.

Nathaniel Schooler
Oh, very much. So. But I think, I think, I mean, I was thinking about it, been thinking about it for years since I sat around that roundtable with IBM, and listen to the editor of wired, and a few sort of top top people. That was in 2015 and the fact is, is that there are, there is a lot of hype, there are going to be a lot of people who are out of work, but at the end of the day, those people would, would potentially be out of work anyway, because they’re not, they’re not learning new skills, they’re not interested in, in furthering their education. Yeah, and that’s, that’s a, that’s a big, big problem.

I mean, data and an analysis of it and using AI (artificial intelligence in business) it is in everything underpins everything, Internet of Things, security, you know, every everything, right? Like, if you think like, the amount of the amount of CDOs, which if people don’t know what that means, it’s a Chief Data Officer or Chief Digital Officer, one or the other, it doesn’t doesn’t actually matter too much, because they have actually almost merge those two, those two things, or at least they should, they should have merged and, and a lot of those have been pretty promoted to the CEO now, because they’ve actually done such a great job as this as the CDO that they were like, well, this guy’s done so much, we’re going to promote him to the CEO.

Daniel Yarmoluk
It is definitely it’s definitely changing and I find it incredibly exciting, you know, one of the things you know, so there is a lot of hype, I think we’re digitizing, you know, small portions of problems, and how do you apply them and I talked to an interesting guy from Deloitte analytics team.

He said that a lot of CEOs are asking him for AI. and he says what do you want to do?

Well, automate and automate suggests, you know, cutting costs cutting humans are cutting processes and cutting costs but realize for the next couple years, you got to maintain your current staff at another staff to add the machine learning models, and then that has to be tweaked and maintain. So that that is a net addition of jobs in the short and medium term as they try to get to automate. So this kind of re distribution of, of job tasks and jobs, I mean, is is is to be questioned.

I think you’re right, there’s there’s opportunity to get educated. And now speaking of disruption or digital infrastructures, I mean, there’s so many certifications you can get online, it begs the question, do you need a computer science degree if you got all these certifications from all these agencies? So it’s all shaken up? And again, what’s the value? Should I drive two hours after work? Or should I take a class online for one 100th of the price?

Take a machine learning course from Stanford, any founded Coursera and these or I can do it for 4000 down the street.

Nathaniel Schooler
it is nice to meet other people and get that extra. You can do that on social and everything. But I think I think from where I’m sitting the education I mean, I’ve been learning Spanish for years.

The first time I really accelerated my Spanish was when I downloaded a Duolingo Spanish, right? And, and if Dulolingo can become the most downloaded education app in the world, they are doing something, right. All it takes is someone with a frickin brain to go, Oh, well, maybe I’ll build one and I’m going to do it for business then they’ve got the next education disruption thing. That it’s not the same as going to a top school or even a good school and actually meeting people face to face.

Daniel Yarmoluk
You’re absolutely right. But when you when you think not, everybody can, it just democratizes that option. So is it right or wrong, it’s different. It’s an option and for the person that wants to do it and has the gumption to do it, they can, they can learn something, whether it’s

I am not a you know, even though I teach part time, I’m not I’m not in the educational theory. I don’t know all that stuff. But, you know, I’m just looking at the options to accessing information. And then that is pretty interesting. I also heard people, you know, there was a really, really interesting TED talk about this gentleman was an educator in in like rural India, and he, his, his office buttered up against an Indian slum. He put a computer terminal in the wall and that, you know, people like a poor children in India would go up to the terminal, and they learned how to speak in English and program and they kind of like with nothing, just a terminal sticking out. So that that tells you with the people that have the ambition to learn what they can learn.

They replicated the experiment through India bus stops. So then the question becomes, we can scale up cell phones, but we can’t scale up humans that are educated to go in all parts of rural India.

So, I say that to think of the problem you want to apply it to, it is not just AI, AI! It’s what are we going to do with it, where we going to put it whenever you’re doing machine learning, or any of these things, it takes a hyper sophisticated person and domain to write these things.

So you got a complex computer problem, you got a complex domain problem and then you have a complex execution problem. So all those things are going to have to be in place to maybe have a shot. So yeah, I’m going to see so many pilots and failed experiments that this is kind of like a revolution.

The futurists that I read say that, you know, 50% of the Fortune 500 are going to be out of business to the next seven to 10 years

Artificial Intelligence in Business is Over Hyped

Nathaniel Schooler
Yeah, yeah, I think they probably will you know, there’s so much hype about artificial intelligence in business and so much and they’re so slow a lot of these companies and people to

Daniel Yarmoluk
My buddies as as well as you in the people you interview are telling me their own internal conference calls for seven hours a day in this large globally matrix organization.

Then you’re gonna have some small digital company as slack and Uber that just just pierces right through it, because they are slow.

So then this gets into where I think the battlefront is made. Where I really think the battle is is this whole cloud agile development sandbox, where whether that be AWS, IBM Azure, people are being able to develop things in the cloud in a risk free, cheap way that’s available to all an experiment.

The more you get up to bat, the more you’re uncomfortable trying new things, you know, some experiments will succeed. So I don’t think the battle is an AI. I personally think the trench warfare the forefront of the warfare is people that doing experiments in the cloud with this cloud, agile development, you know, micro service world, that’s where I firmly believe, like the, you know, first ending first round, whatever you want to call it is, that’s where it’s at.

Nathaniel Schooler
Yeah, because they already have loads of the infrastructure already built. They’ve got they’ve got a lot of the data bill, they’ve got a lot of the questions and answers built, they’ve got, they got loads of stuff built, and you can do it in 200 languages. It’s going to translate it for you. Just like that, right

Daniel Yarmoluk
I know, guys that I am talking to across the pond. I mean, and that are well travelled. I mean, just the idea of the whole Google earbuds we saw six months ago and translating.

This is this is a marvelous thing now that if we can think of a world where they were in China, and communicating independently with people through I don’t care phones, apps, you know, and just communicating in a human way and that just seems super exciting to me.

If you would like to listen to my episode with Founder of the CDO Summit and CDO Club David Mathison here is the link to the future of big data. data