So today I’m joined by Dan Yarmoluk who’s an adjunct professor in artificial intelligence machine learning and he loves a bit of IoT as well as industrial internet of things kind of stuff and Craig Truempi and he’s into AI, Big Data and machine learning. The same as Dan really and he’s a director of AI and IoT digital ecosystems. So we just talked about latest updates from think 2019 what’s kind of happening in machine learning, AI hype…
You can also learn a lot more about AI/IoT from the IBM Maximo blog, in particular the key area of asset management.
Nathaniel Schooler 1:01
And also we talk about how to set an early warning system for your mother in law.
Nathaniel Schooler 1:10
So I’m quite excited to speak to you guys. It’s nice to meet you Craig. Nice to see you again. Dan. So what’s been what’s been happening over at overthink in San Francisco today?
Daniel Yarmoluk 1:26
Well, I mean, I think there’s hundreds and hundreds of sessions of thought leadership of, of innovation, you know, but really, I get the feeling like the the dialogue overall is this multi cloud hybrid cloud strategy and the migration path right nation path can be an emphasis on cloud or it can be an emphasis on the AI machine learning for the future.
Daniel Yarmoluk 2:00
But getting more tangible, more more hands on coding development workshops. I mean, Craig, who’s a double engineer from industry for years, went through a code development little thing. We’re building Chabot in an hour.
Daniel Yarmoluk 2:15
So just hands on work. It was a chat bot and in 15 minutes. Yeah, it was amazing. It was hands on. And it was using their model to build yourself a chat bot. And so I haven’t I haven’t programmed or coded in probably 20 years really. And although I didn’t step that far away from it, but you know, we got in and we did it and it worked. So it was, it was it was pretty cool. Wow. So that’s actually already kind of stacked up with questions and answers. So if I come in there, and I’m in you know, and I ask a stupid question, or i or i swear or I do something, it’s going to come back to me with an answer. Right, right. So you build you build the questions that you think somebody might ask and you
Craig Truempi 3:00
Then lay out the, the, you know the tree and how the answers might play out. But the interesting thing is, when you lay out your questions, you don’t have to have all the questions there. It has AI or Watson built into it, where he will go, or she, whatever Watson is and will develop kind of that greatness. So if somebody asks a question, similar, but not exactly, it’ll figure it out, right? That’s pretty cool.
Daniel Yarmoluk 3:26
Yeah, it is pretty cool. So even to the last workshop we just got out of five minutes ago was cloud migration strategies, kind of, they have a Chatbot to kind of chat with you to discover, which workloads do you want to shift to the cloud. So AI begets AI from the cloud journey to enable your AI personalized initiatives, meaning I want predictive analytics in the plant but I don’t know what to ship to the cloud.
Daniel Yarmoluk 3:53
So therefore, this cloud kind of automation says this workload on this server so having therefore you might want to share. That the cloud so it’s, it’s it’s all converging it’s this convergence and collaboration, sharing and one thing’s for sure is when we look at this large partnership network we all got to leverage each other’s tools it’s the integration of all this massive amounts of very siloed IT systems and getting it together and that that integration is not is not easy but is it’s getting better.
Daniel Yarmoluk 4:29
I think I’m more refreshed after 365 days from last year to where the industry is getting. I like the way the dialogue is moving. I mean, at the end of the day, I mean I really feel like I’ve been sprinting for three years I can’t catch my breath but can absorb as you know, with all the guests that you have. It just keeps coming and coming and it’s just super exciting.
Craig Truempi 4:54
You know, it isn’t it’s getting more real, like there was a number of customer presentations kind of talking about how they are using this technology. And then I come from the IIoT side of things, your operations side and that.
Craig Truempi 5:08
There are heavy IT people here, but you see them learning our vocabulary and us learning there’s and the other people that I met were really asking some very good questions around IoT and I assume they plant side people and they were IT people and so that’s that’s refreshing that’s kind of confirming the convergence is is is occurring.
Nathaniel Schooler 5:31
Tight i think that’s a bit of quite a bit of misunderstanding though around how IoT works and around how machine learning works and AI, I was talking to a friend of mine today. He does a lot of stuff for himself, you know, he is a programmer but he but he actually builds you know, Internet of Things. So he’s like, Well, how do I set a because he’s got a he’s got a girlfriend So, you know, how do you how do you set a warning system for your girlfriends mother?you
Nathaniel Schooler 6:05
It’s quite a simple process, isn’t it? Like all of this is actually quite simple. He broke it down for me. And before that, I was quite sort of confused about it. But in essence is quite simple. But I think a lot of people just add complexity to everything. And they use fancy words that, you know, confuse people. And it’s like, Look, if I’m gonna if I’m going to set a system so that it you know, I tell Alexa to organize it for me. And then my mother in law arrives and the lights go off and the curtains shut automatically. actually setting that up is quite simple guys. And it really
Daniel Yarmoluk 6:44
Yeah, yeah, it’s become a lot easier. It’s become a lot easier. I mean, in this
Nathaniel Schooler 6:50
field when your mother in law comes around.
Daniel Yarmoluk 6:53
Well, Craig, and Ijust said we could build a chat bot on text to respond.
Daniel Yarmoluk 7:00
On to our wives, you know, complaints when we’re travelling. So but but we got to make sure that we ask the right questions at the chat bot so we don’t screw up and get caught. We have loving responses just auto generated at the right time. I mean, I think Dude, I think we should do it see if we can pull it off.
Daniel Yarmoluk 7:25
So you’re absolutely right. I mean and the are the possible the question that I always come back and the reality is they have to know the data set and the in the domain which is you know, you know, is Craig right here knowing just knowing that data because the the model the model the model the tool. The vessel, you know, what is the output is that beneficial and he’s rejected a lot of machine learning companies because the insights he generates is 30, 60, 90 days in advance, like, you know, with your father on the factory floor and the machine learning guys we’re getting one or two days in advance and this is not going to work at the factory.
Daniel Yarmoluk 8:01
So this is a collaborative effort of technologists and professionals and they’re open and THINK is here to discover and learn, and even as communicators or integrators, it’s confusing. So it’s confusing to a lot of people, because there’s so much advancements coming up every direction, and you look no further than the auto industry to see the insanity or mash up of these technologies. That was really inspirational at the IoT with Dr. Youssef, the IoT Watson guy, and yet the Mercedes guy last year, if you remember, I don’t know if you’re there at the Jaguar Land Rover. So mobile entertainment, battle, electrification, battle, autonomous vehicles, people’s millennials shape you know, choices changing and then industry 4.0 on the factory floor and collaboration to elevate everything going on. And communication is just immense.
Daniel Yarmoluk 8:55
It’s just like insane and they were talking about Mercedes and BMW have a joint partnership for electrification and like, a three or something or I think it was called right, you know, where they’re going to share this autonomous vision and electrification vision, so it’s like it’s just super exciting.
Daniel Yarmoluk 9:15
I just hope to learn from these smart guys and apply it in our own little way in a meaningful way and I think will we will be you know, we’re not going to be unemployed anytime soon. I think with all this Yeah, someone’s we keep up with learning. I mean, that’s what in essence that’s what this is all about, isn’t it?
Nathaniel Schooler 9:34
It’s about just understanding that you know, you haven’t just go out University you don’t know everything you know, and you need you need you need to learn and wherever you are in your career if you’re not learning you’re dying. I’ve been saying this for years but no one really took me seriously until all this stuff sort of started happening you know.
Daniel Yarmoluk 9:53
We can we can kind of kind of go to that point you know, Ginnie you know, at the keynote you know. And I think this is important in a global context. And the visa issues and the technical skills gaps, was talking about a lady who stayed at home, she was like going to be getting a PhD and rocket science stayed at home for 20 years.
Daniel Yarmoluk 10:13
And then she called it a real internship coming back into the workplace with certifications, not just old school degrees, and an army guy that a young, you know, intellectual property type, like looking at at all. And I thought that was inspiring, right? Because you and I both went through that journey of going back in and trying to reboot what we’re dealing with what we know, and that’s what it’s going to take.
Daniel Yarmoluk 10:38
So I thought that was very meaningful, but diversity of thought and opinion, but also how are we going to get there with these shortages? Because there’s just shortages of these skilled qualified people. Oh, yeah. Learning and that’s what we’re trying to do. I mean, also in my little experience with Craig I mean, I’ve been taking them these data science conferences, not because I know but to show them a tool set.
Daniel Yarmoluk 11:00
That he can employ. And then this beautiful things have happened with stuff like service functions. And so it’s been it’s been it’s been fun. I just want to get to work with all this stuff. Now
Nathaniel Schooler 11:13
I’ll talk about what’s your what’s your take, then Craig? What, what, what excites you the most of all of this?
Craig Truempi 11:19
Well, you know, a couple things that that Dan mentioned. One I wanted to just drill down a little bit farther was on the data side of things. So there’s a quote in earlier conference that was data makes the difference between average AI and great AI.
Craig Truempi 11:34
And so, you know, when you’re looking at doing these magnificent, magnificent things with AI and you’re looking to pull the data that’s available to you, is that average data is that going to get you the great AI?
Craig Truempi 11:45
Do you have to think hard about adding measurements that are meaningful to solve the problems that you’re looking at? And so I it was refreshing to hear the people that are in that AI circle. #
Craig Truempi 11:56
Understanding that data, the data stream makes the difference and so that was rich other piece that was kind of exciting for me, we’re doing a lot of things with the sensors in the IoT space, but it was the acknowledgement that it’s the AI plus IoT that really gains the value, right?
Craig Truempi 12:14
It’s, it’s about that insight that the AI has the ability to process or analyze this data in a meaningful and textual way in automate some of that analysis so we can get to the action where the value is created. And so many times I think IoT can end at the alert stage in the alert.
Craig Truempi 12:34
The alert is and what we’ve what we’ve seen an industry it’s not enough to get the action to tie it out with artificial intelligence into natural language and and integrate it to existing people systems like the IBM maxima work order system.
Nathaniel Schooler 12:51
What does that do then? Craig, sorry to interrupt you there?
Craig Truempi 12:55
Sure. So, with that, you know, the IBM Maximo system is is a maintenance man management system, a work order system for maintenance. And they say, you know, it’s not in Maximo, the work doesn’t get done. So by automating this alert to integrate with Maximo and generate that automated work order that work has a better chance of getting recognized, getting listed, eliminating some of the administrative tasks that can be the obstacle or barrier delay and in get that work done, and actually get the value in business results.
Nathaniel Schooler 13:29
Right. Right. It makes a lot of sense. I mean, I think there’s a lot of things that can be automated, that can save us a lot of work and a lot of time, a lot of effort.
Nathaniel Schooler 13:37
And, you know, as far as I can see, I think, I think it’s, it is an exciting time right now. But I think also for a lot of people, they’re quite worried and I don’t think they need to worry because, you know, the companies that have this technology, IBM and all the others because there are many others. IBM isn’t just the only one it is a leader. There’s no doubt, but you know you’ve got a lot of companies that are using it.
Nathaniel Schooler 14:03
And I don’t think they need to worry because those companies don’t want the global economy to crash. Because, you know, the more people that are out of work, the less money is available, then everything just goes wrong. Right? consumerization would in essence die. I mean, that’s could be a byproduct, right?
Craig Truempi 14:24
Right. But we, you know, we see even in our businesses that that we have in manufacturing, that hiring talent is is very difficult today. So, the automation of some of that is it’s, it’s ripe to just keep the businesses running.
Craig Truempi 14:41
And you see the demographics of, you know, that tribal knowledge exiting these companies. And so can I fill you know, fill that growing gap. It’s not about replacing, you know, people that are willing and interested to work. It’s replacing the demographic of people that are leaving and making it easier and more tech for the people coming in that want in, accept the technology.
Craig Truempi 15:05
So I think there’s timing that is, you know, making this very, very appropriate.
Nathaniel Schooler 15:12
That’s a very good explanation. Thank you, Craig. I appreciate that. It’s been great talking to you guys. And I hope you know, we’ll speak soon but I’m going to get get going because I it’s kind of getting getting on a bit here and I want to cut this up and post it. Very cool. See you guys. Alright,
thanks so much for listening. Please subscribe and wherever you prefer, share with your friends. And if you enjoyed the show, drop us a review on iTunes or wherever you listen.
Listen to Dan talk about AI IoT and Machine Learning here too!