Season 1 | Episode 7

Ben Garvey 

That there’s a lot there. 

There’s a whole lot to juice play through that. 

Jen Laplante 

There is. 

Ben Garvey 

I mean, I mean I, you know, you just when you start with, you know, sociology. Yeah. And and human behaviors and human science. And then you you dive into data from that you know, you you somehow managed to make that leap from the sociological study and behaviors and data into entrepreneurialism, into business, into telling that story, I mean. AI AI, machine learning in general. Really, really exciting for techies and for folks focused on it. But how do you how do you know you’re in a pretty unique position in it? You’re successfully translating the tech needs the tech objectives to management and to leaders. How do you get people excited? Because I mean, technical people get excited really quickly about AI and data in general, but it’s generally a kind of a dirty topic. I mean, in many boardrooms, right as I how do you bring? Boards of directors and management up to speed and get them excited about this. 

Jen Laplante 

I think the. Biggest thing well lately, to be honest. You know, we really seen ChatGPT take hold and generative AI, so it’s a bit of a different news story than it was back in 2019 when I tried to sit down with companies to explain what artificial intelligence is and how machine learning might be able to help. I think that the biggest thing you know today is really around the optimization and opportunity that companies have to adopt new technology. There is a piece too, where some companies want to do it just because they think they want to air quotes. Do AI right and that’s one and of itself. You have people you know are coming to the table and saying I want to do. Something the biggest challenge that was stepping back and saying what’s the problem you’re trying to solve and we really have to nail down what that problem is. But as soon as the company says that we want to do some form of AI, it’s or it opens a door on the other side. I think that the challenge that we have is there’s lots of companies out there who could really benefit and they just don’t know where to start. And so that’s the other the other piece that you know we can spend loads of time talking about how do you actually get technology adoption to grow Atlanta, Canada? You know, we’re already a bit behind on technology adoption in general. The ocean sector tends to be based in Atlanta. Canada, Canada in general is still slower on a digital adoption than the rest of the world. So we have a lot of catch up to do and hopefully it’s kind of like those countries in the world that kind almost skip technologies. Maybe we can do some skipping of technologies and have a broader adoption here too. 

Ben Garvey 

Cool. That certainly resonates. You alluded to it a little bit in the kind of the early days, there’s a little bit of a hype versus reality. Where are we on that kind of curve now of? You know, there was a is going to solve everything and and you also kind of referenced you know understanding what the problem is you’re gonna solve right there. There’s this in my mind. I see attention between, you know, domain expertise, you know the the people who are, who are expert at solving these particular problems in this domain. 

Speaker 

Hmm. 

Ben Garvey 

But without AI and then bringing in tool, you know with, with a kind of a big fanfare of saying, OK, this tool is gonna make your job much better, much easier. Where where are we on that kind of scale now with any new technology, any new advancement? There’s a there’s a bit of a rebound. Times. Where do? You see that? 

Jen Laplante 

There’s it’s kind of a tricky hard part to pinpoint, so small and medium sized enterprises I think are a little bit slower on. The adoption of this space. Yeah, they simply don’t have capacity. The money, even the change management aspects associated with developing and implementing some kind of new technology that is AI as a tool, is is helping them as hard larger enterprises they think are a lot further along because they have either the budget, they have somebody who’s focused or job or you know it’s on their objectives for the years to understand how we can kind of adopt new technology. So they tend to be ahead. I think that. The whole piece though for me is is more around thinking about. Go back to the problem that companies are trying to solve, and there’s an element here of the difference between build versus buy and if the company needs to be building it versus buying it. And I was at a board meeting last night, for I am on a couple of different boards and this was a not-for-profit. And someone said, well, can we do some AI for that? And I was like, but what’s the problem you’re trying to solve? And it goes back to what are you trying to do? And so when I think of the companies that want to do something, they think necessarily they want to go create AI and they don’t realize back to your. Mean you need first the data you need, don’t domain expertise that can help you understand what that data is and what it’s saying and how to label it, but I think they also don’t always realize now you need software developers and you might need a cloud engineer and you’ll have to think about the front end and the back end. And how does this whatever this tool is, it gets deployed and you somehow, how do you actually use these models going forward? And so then the build part is still, I think a really small piece of the overall number of companies out there. And BCG has this great graphic that shows innovation and it’s. Have like you know the time horizon and the opportunity and it kind of is taking if you have something that’s core part of your business you can innovate just outside of the core part. There might be some adjacent types of services and technology or products that you can use, but if it’s so far outside of the core part of your business and I see a lot of companies you think they want to adopt AI but they don’t like they might have an outsourced person who images. Their PCs right? Like and, they’re not going to be the people who go and create some really cool novel AI. 

Ben Garvey 

Right. 

Jen Laplante 

Solution. But then I think it gets into the whole piece of buying something and that just opens a whole other can of worms because some of the companies out there who are doing really cool AI solutions are still so new that if your company do you necessarily want to put all your eggs in one basket and are you going to rely on this company and hope they’re going to be around forever and that’s a whole other challenge. And then I think too, the transferability, you know, I’ve seen enough aibs out there to then question is this a company? Worth somebody? Actually involving themselves with because do you trust the reliability of what they’re producing? You know, I’ve heard so many things around. Well, it’s 85% accurate, 95% accurate. If you’re doing diagnosis of radiology, I don’t know if I want my doctor using something that’s 85% accurate and if I’m doing corrosion on underwater pipe, is 85% accurate enough? Well, maybe it is, but then I’m not actually sure what the 85% accurate is of. So we have this whole slew of these companies out there who want to figure out what they can do, but they just, they don’t know if they where they fit in the buy versus buy. World. 

Ben Garvey 

Yeah, it is. It is certainly, you know, back to 85% certainty that AI and machine learning in general make mistakes, right? That’s part of the whole thing. That’s part of the learning. So yeah, knowing where, when, how to, how to account for that is part of the magic, isn’t it? It’s it’s both part of the excitement and part. Of the terror of the whole thing. 

Jen Laplante 

Yes. 

Ben Garvey 

That’s kinda cool. Uh, so in in Ocean’s tech in general. Well, in Ocean’s tech specifically, where are we seeing the the value add today for for AI, where is it? Where are we seeing the practical application? Where is it being used? Where are we seeing kind of the the rubber? Hit the road. 

Jen Laplante 

I think the biggest pieces we’re seeing is around data volumes when data volumes are so sufficient that you need some kind of machine to help go and sort through. And identify or from a prediction standpoint. And so we see this, for example, in aquaculture, where there’s significant amounts of data around water temperature and salinity and making predictions around the quality of water and perhaps maybe algae or other types of negative. Events. Um, we’re seeing it too. Around where you know there’s too much information for a human to do something, so there’s elements around sorting through images and I and and identifying patterns. Even some aspects associated with fatigue. So you have. 

Jen Laplante 

Lots of information around a worker and a worker’s performance. And can you start now identifying where they may be getting too tired by pulling multiple forms of data together? Everything from maybe the whoop that you’re wearing to images around your behavior of how you’re walking around a vessel or a platform or some other site that might be dangerous. One of the areas that I think we’re really trying to push to is around. Um. Autonomy on the ocean. So um marine surface vessels that can operate autonomously. You know the Teslas of the ocean? Maybe Tesla bad example. We should use that anymore. But you know the autonomous systems on the surface, like there’s significant opportunity for us to start thinking about collecting more information in the ocean. 

Jen Laplante 

That’s really valuable. And when you look, you know, we’re based here in Halifax and we were supposed to have the worst hurricane season on record. Fortunately, we weren’t slammed, but the rest of. The Atlantic Coast. Was we need more data about what’s happening out there and so if we can go and send more vessels out there, that’s great to collect data. So there’s lots of companies working on that. I think that you know on on the marine side associated. If transportation and logistics we’re seeing way more collaboration, which then results in data sharing which can help optimize and understand what’s happening with marine vessels. And so if you’re bringing vessels from across the ocean, better prediction better, you know, use and reduction in greenhouse gas emissions. So I think it’s a lot of it is really just people coming up with the bigger picture problems and then trying to find partners where they can either share data or share skill sets to try to find some neat solutions. 

Ben Garvey 

Yeah, I mean, I would think that there’s huge data sets out there in, in environmental science already and and lots of complexity there that that I’m I’m sure is being used. But yeah, I’m hearing lots of of discussion about, you know, swarms of of data gathering boys and uh, you know, sometimes that that triggers. Kind of discussions about the darker side of of AI. Perhaps you know the some of the the objections that might come up. How do you know some of the things like, well, personal personal data security, environmental impacts of of, you know, mass computing, um and social engineering from data, these sorts of things. How do you balance those sort of they would be objections? How do they get balanced among the kind of the thrust to use the the the tool for good? 

Jen Laplante 

Yeah, I think, well, I’m fortunate that in the ocean ocean sector, we’re typically not looking at a whole lot of personal information. And we’re not making independent decisions. So you know the examples I have to use when I’m giving talks talking about the ethics of AI. Like if I were a judge to Sir, you know, deciding somebody sentencing or, you know, a bank deciding someone’s likely to get a loan or, you know, a treatment or decision, whether or not someone should be receiving a vaccination based on some kind of formulation. I think those are the ones where the ethical concerns are quite strong and quite scary and I definitely am very worried that we don’t have enough knowledge and information within our government departments for them to make those decisions because things are so new and people don’t really understand what their risk is and you can easily trust a vendor as they’re selling you something that’s a great book. I recommend is a AI snake oil, and there’s a lot of things around it. 

Jen Laplante 

It’s really good. I’ve really good book. I can recommend so many good books, but that’s a really good one I just finished. But there’s some pieces around the ethical side of that. I think on the ocean side, you know. Um, the the biggest piece that we kind of run into a challenge with really is, is is around the data sharing whether it’s something is private data from a company and they don’t want to necessarily share it with someone else. We are in the realm where we need to trust AI a lot more, which means we kind of want models to be available open source where someone or someone can validate and understand and trust. What you’ve trained will now the sudden you have to make your data publicly available and so that might be more of the risk that. We’ve seen on the computer and the environmental side, I mean really, that’s generative AI. And so it’s the fact that you have these large compute clusters that are required to train these massive frontier models and these LLNS and create generative AI. And those are the things that are requiring all the NVIDIA chips and the net new data centres and I think it’s this piece where we have to be really mindful if you’re doing a a. ChatGPT search it’s. 10 times more energy than it is for a Google search. And so it’s, you know, we’re creating this demand and if you look at Microsoft. To you know, they’re reduction of emissions was expected to go down in 2023 and year over year. Now they’re up like they’re up significantly. And so they’re not making reductions. And so when we think about AI, sometimes it gets painted with the bad brush, right? If I’m going to use images to identify something like, I’m going to use corrosion underwater, so that’s really important. If you have underwater pipes or let’s say you’re inspecting our Internet lines for example, that are underwater and subsea. Um that cross from 1 country to another? You know those are images and footage that can be analyzed, probably with some on premises compute, sometimes even a laptop. If you have a strong enough laptop, we’re not, you know, causing catastrophic harm. But the big generative AI pieces, which are neat to see in somewhat useful in some respects, are the ones I think that are having way more damage. 

Ben Garvey

Interesting. And so in a global context. And we’ve been talking sort of locally Atlantic Canada, Ocean tech, where in, in a global context do we do we land in the kind of in the AI app in the application of AI in the blue tech world, are we are we leaders? Are we?

Jen Laplante 

That’s a great question. I you know, we’re working on figuring that out because in the ocean space and AI, and that’s one of the things that I think we ever real opportun. 

Ben Garvey 

Competitive. 

Jen Laplante 

Right now, there aren’t any global leaders. Before so, Canada is one of the leaders in in artificial intelligence. Back in 2018, when Canada created the Pan Canadian AI strategy, you know, funding was originally provided to a group called Sifar, which then funds three of the major research institutes in Canada. One of them called Mila, the other ones vector, and one of these, Amy. And so they’re in Edmonton and Toronto and Montreal, and these are the hubs that have the greatest strength. And so in 2018, we were some of the leaders in the world in AI and I think you know, that’s by the government decided to invest so significantly. In this, unfortunately, that didn’t always translate into actual adoption. So there’s one thing to be a leader in, you know, training new models and writing papers and publishing them and having great researchers to, you know, having some small and medium sized businesses adopt the technology. There’s a huge disconnect, and ocean sector tends to be small and medium sized businesses for a lot of the part right. 

Ben Garvey 

You could. Yeah. 

Jen Laplante 

And so there’s this piece where I think you know. Canada could be a really strong leader in Ocean AI and really setting up a platform when we look around the world and we go to conferences or talk to other groups, there isn’t a whole lot of conversation around AI. It’s starting to happen. Maybe a little in Portugal and a couple of other places, but the conversation is still pretty new. And so there’s there is this opportunity for us to really push and show how we can leverage both our strength from an academic side and then bring that over into small, medium and large businesses in the sector. 

Ben Garvey 

Cool. That’s that’s great to hear. And would you say that the weight is kind of 5050 on the development versus deployment side or is it? There’s a 6040 or we doing more R and. D or more practical? 

Jen Laplante 

Oh, we do way more, are indeed, but I think we have to be mindful. So I mean, there’s a lot of research out there around AI that you know up to probably only 25% of attempts for projects will actually work, right. 

Ben Garvey 

But great, yes. 

Jen Laplante 

So you don’t know and I. Think this is so? This is the one thing we’re also seeing is you could ever. Really large enterprise. Who’s deciding to do some kind of AI solution? They’re going to make themselves and deploy or purchase from. Where the problem you have though is you have an annual plan. You have a budget. You have your cap and your OpEx mapped out and if you can’t demonstrate a really good business case in ROI, it might be easier for us to go build this new building, buy a new vehicle implant, implement a new ERP system. And so you have this challenge where even some bigger companies that could be deploying more frequently because there’s still some ambiguity around the outcomes, they struggle themselves in trying to prioritize deploying some of this. 

Ben Garvey 

Sure. Yeah. I mean, it’s it all comes back to ROI and and demonstrable our right you have to be able to describe it and show it if they’re in. 

Jen Laplante 

Yeah. Yeah. 

Ben Garvey 

To uh, back in a little bit higher up than the AI world, which is pretty high, but. The Ocean Supercluster itself, I mean you’ve been involved in it pretty heavily for a while now. Just give give me a little background on how how that supercluster is is fostering innovation in the industry across the country now. What are you seeing there? And and where’s the? Where’s the magic? 

Jen Laplante 

Yeah. I think I’ll take a step back really quickly. So the superclusters were a program developed by the federal government in 2018, Ish, and so there are five different clusters across the country, focused on different domains. There’s manufacturing, for example, in a based heavily out of Ontario, there’s a proteins 1 based out of the out of the prairies. But you know we’re I think the thing about us is were not-for-profit member LED organization, we’ve been given funding from I said the Government of Canada to invest in collaboration project. And I think the things that’s unique about what we’re doing that’s different is our projects are collaboration. So unlike, maybe we’re someone might be going to enter C, Iraq, who might be able to support them on the development of a thing. It tends to primarily be a lot more one company and their support there or even if they’re getting funding through a coil or or, you know, fed Dev or one of the other provincial or regional development agencies, you know, ours is really much on the collaboration side of things. 

Ben Garvey 

Yeah. 

Jen Laplante 

And so I think that’s one of the really great things that we’re really seeing is companies coming together. You know, we’re in the second five-year mandate. We’re a year and a half into it and so I’ve been here for a year and a half and what I’m. Thing is, we have a 700 member strong group. We’re seeing really good collaborations. We see a lot more people coming to the table saying I need this and so the strength that I’ve observed is is companies, you know, walking in and trying to build a relationship with someone to create it together. And I think there’s strength in numbers that if you’re bringing your customers to the table, you’re bringing to somebody within your supply chain to the table. And we’re investing with you in innovation. We’re helping reduce a bit of that financial risk. We’re not giving you technical expertise, but we’re at least how trying to help you give a bit more incentive for you and your partners to work together, which I think then results are really good IP, really good products and then hopefully you know. Greater relationships that continue to grow over the years and they don’t necessarily need our investment because they’re working so well together. They found new opportunities and off they go. 

Ben Garvey 

Right. Are you seeing the commercial successes evolving? 

Jen Laplante 

We are definitely and I think the the biggest challenge is innovation takes time. And so it’s really interesting. I was just reading a feedback survey from some of our members and someone said, you know, it’s hard when you operate on a funding cycle because success measures to say the day after your funding ends, you’re not going to go turn around and say now how many jobs did you create? 

Ben Garvey 

Go. 

Jen Laplante 

They take time, but the companies that we’ve been investing with really have seen some really good growth. Their products are developing or they’re continuing to have follow on investment and development. Um and I I think that you know, if we look at what the ocean sector looks like in Canada and Atlanta, Canada, 7 years ago, we’re fundamentally a different country. We have way more companies. We have way more people employed here. The growth has been really strong. 

Ben Garvey 

Yeah, I certainly see that too. I mean, we’re we’re involved in it on a daily basis and it’s phenomenal to. See the growth, but what? It’s it’s great, it’s it’s started the innovation. We’re seeing some successes coming out of it. What is what are the obstacles that these companies are facing when they kind of mature a little bit and step out into the into the world. What do you see as the? You’re this proverbial value of death that people have to get across. 

Jen Laplante 

I have so many valleys of death. 

Ben Garvey 

You know, yeah. 

Jen Laplante 

I mean, we see some one of the great things I think the supercluster did was the ocean startup project and they helped fund and inside that. And so we’ve seen it over 100 startups come through. So we see them and other companies will kind of come through and create their product or service and getting those first customers is the struggle. Sometimes there’s the regulatory side that’s a bit of a hurdle. So um for example on. And on the AI side of things, we have a couple of really neat projects around environment electronic monitoring. So in BC, for example, you have to have cameras on your vessels that capture images or capture video footage of you as you’re doing catch. And the whole idea is if DFO were others one or then go and review what you actually caught to ensure you’re not catching over your quota, you’re not catching and keeping things you shouldn’t. They’ll review the footage. So now there’s some companies archipelago on deck that have created these really great mill models that have been built out to be able to analyze the software and clean. And process it. It’s a whole other story. Getting DFO comfortable with AI. So it’s a great solution for something that’s been mandated. You need to have cameras on. You need to be capturing. Someone needs to review the data and that’s a challenge. We’ve even seen this on the side where, you know, under underwater images around fish, at hydro dams, even entitled turbines like the analysis and processing of data and making some government bodies comfortable with these really great solutions that can enable, you know, whether it’s great renewable energy development or just optimization of. Of fishing like these are the things that I think are are there challenge some companies face. I think the other thing we see too is some people are really some companies are really great solutions that government as their customer in general and the buying cycle for selling to government is really hard. 

Ben Garvey 

Hmm. 

Jen Laplante 

I’m sure you’ve heard you know that we constantly hear that the dual use where you have a civilian or a civil application as well as a military defence. 

Ben Garvey 

Yeah. 

Jen Laplante 

Application and those are fascinating, and I think, you know, we’re really interested in something that could have multiple markets and really grow, but even still it goes back to the whole problem that you know, you’re probably selling to a large prime like 1A defence contractor, you’re selling to a large company which in and of itself takes time. 

Ben Garvey 

Redstone. 

Jen Laplante 

So the selling the selling part is really hard. 

Ben Garvey 

And that’s that’s across the board. It really doesn’t matter if you’re in blue tech or anything right then. 

Jen Laplante 

It is, yes, everywhere. Yep. 

Ben Garvey 

The aerospace defence doesn’t matter where you are unless it seems like you’re in consumer products. That’s it. You know the you get the whole yeah resolution. 

Jen Laplante 

Well, that’s a whole other some SAP solution. Yeah. 

Ben Garvey 

Exactly, yeah. Is are we seeing it used to be when when we would talk about, you know, innovation in the blue tech world, there would be no money and financing would was always the the, you know, Oh my God. In Atlantic Canada, trying to get a blue tech solution, finance was a problem, and I. Think that’s changed? Or is evolving are you seeing the same thing now? 

Jen Laplante 

I think it’s a hitter, miss. I think there’s a couple pieces. One is there’s a lot of non dilutive funding out there for companies and I think that’s good and bad cause I think sometimes you end up supporting companies that. Should have just dived. 

Ben Garvey 

It should have died Def def. 

Yeah. 

Jen Laplante 

And so then they continually chase and chase and chase and they. Going um on, you know, in terms of the actual access to some kind of capital, I think there’s a couple pieces here. One is climate tech and ocean tech. Confused climate tech. A different mandate. So you know some of those ocean spaces are net new for a lot of different types of funders, whether it’s VC or family offices or others. It’s a real challenge, but I also think we’re in a complete downward spiral right now for access to any kind of funding and the amount of funds have been deployed in the past year and a half has been, yeah, dismal so. It’s it is a real struggle. I mean, we were just talking earlier before we started at two different companies that just closed rounds this year, which is impressive. But there those are few and far between, and they’re not huge rounds. 

Ben Garvey 

And they’re not big rounds either. 

Just like, yeah. 

Jen Laplante 

And so I I think there’s this piece where, you know, companies are going to struggle. But I also think that means the ones that survive are probably going to get, you know, they’re going to, they’re going to be scrappy, they’re going to hopefully probably bootstrap a lot. They’re going to retain equity, which isn’t a bad thing at all, and so hopefully we’ll see some of those ones that really kind of come together and. Can grow and and it’s it’s. It’s tough, like there’s been lots of conversations around a Canadian fund, specifically on ocean, and there’s been a lot of false starts and it would be lovely to see something like that get kicked off. There’s a few now in the states, but they’re really there’s nothing’s happened yet for that specifically as as an area of focus in Canada. 

Ben Garvey 

Are we seeing those US funds come across the border? 

Jen Laplante 

We definitely are. There’s a couple. One is propeller. They’ve invested in their, you know, there’s a company that we’ve got a project with around car is essentially carbon in a monitoring using satellite images and. Answers. There’s another 1S2G the they’ve done some really great investments, you know, real data which is around underwater equipment, there’s and even EPS ventures, which is out of Singapore. So EPS and the mass of shipping company they’ve invested in GIT coatings and in Volti which is a kinetic energy company startup. 

Ben Garvey 

Yeah. 

Jen Laplante 

So we we we’ve seen some good funds come into the region which is great. 

Jen Laplante 

I think we have to be really careful. We don’t scare it away and we have to be comfortable with. Having some of that investment because some of these companies, you know their capital intensive like you need to have money to get going and if you can’t get it in Canada, having to go somewhere else’s key, I mean it would be great though to have the Canadian investment. 

Ben Garvey 

Oh yeah. 

Jen Laplante 

So we don’t have the risk that those companies actually go elsewhere or have decision making in other jurisdictions. 

Ben Garvey 

Sure, sure. Or or have the demanda yanked from the through whatever cross-border challenges they’ve faced. 

Jen Laplante 

Yeah, absolutely. 

Ben Garvey 

Yeah, it’s always a risk. Is it every see? Thing you mentioned Singapore, we’re seeing buying elsewhere in the world coming to Atlantic Canada to try and grow this. 

Jen Laplante 

Absolutely. I mean we’re. I just was. With people from the British High Commission in the UK. We’re seeing real interest in offshore wind in, you know, decarbonization of everything from fleet. So we’re looking at, you know, hydrogen of itself is a is a contentious topic these days, but we’re still looking at methanol and ammonia and other alternative fuels in the electrification of vessels and fleet. So we see groups from like the Catapult UK group that’s focused on that have been here a lot of relationships have happened through Brazil, Portugal and. Spain, France. Super interested. Even, you know, we were with folks from Norway last week too. There. So there’s a huge interest in information sharing. Understanding that you know, we don’t have to redo the exact same thing that we should be able to. 

Ben Garvey 

Sure. 

Jen Laplante 

Find opportunities to collaborate. I think we’ve also seen Iceland as a small population, but an incredible country in terms of its own focus on fishing and you know, they’ve gotten to the point now where they’ve, you know, the byproducts coming out of fish are now more valuable than the. Itself. And so I think it’s the pieces where we need to start changing our mindset about what we can learn from other jurisdictions that have been really successful and kind of adopt the mindset. 

Ben Garvey 

And is there an open door to that here? 

Jen Laplante 

Absolutely. I think I think there is and I think you have to be careful because if you’re a company and your focus is on, you know, making payroll. Yeah, it’s hard. 

Ben Garvey 

Yeah, yeah. 

Jen Laplante 

You’re focused on that and so it’s the increased conversations with those. Who support the comma? 

Ben Garvey 

Yeah. 

Jen Laplante 

80 and also trying to find net new ways to help support and fund those companies, and so one of the things you know, not back to the supercluster is would be all and end all cause I know we’re definitely not. But there’s this piece where, you know, we just some interesting stuff with innovate UK and we did a joint call for proposals for investing in projects which I think is the stuff that needs to happen too, because it opened the doors for Canadian entities or companies to work with British or UK based. Companies. And they did, you know, collaborative innovation. They worked together and creating new things, and that’s the stuff that I think is really valuable for helping expand the customer base, expand the partnerships outside of Canada and if we can both both countries can toss money at companies to help them do new things, that can be sometimes way more impactful. 

Ben Garvey 

That’s that’s a great kind of segue talking about reaching payroll and then? And building companies to to do this collaboration we, you know, as a as a knowledge based company ourselves we have to be very careful about how we hire the the skill set that we bring in and planning for growth planning for you know tech growth as well. 

Jen Laplante 

Yeah. 

Ben Garvey 

How? How is that a barrier for the for the AI and blue tech industry in our region, we finding the resources, the training, the people that there are we seeing the PEO? Here. 

Jen Laplante 

Yeah, I think there’s a mix. So there’s this balance where you know, if people don’t always know what they’re looking for and so on. The API side specifically, I think that’s a bit of the gap where there is a lot of really great talent. So Munn has a great AI program. You know St. Marys and Dalhousie and UN beat like there’s regionally, we have great students graduating who have a lot of expertise. I think the thing that’s most fascinating is most of them are international. I think that’s going to change in the next little while with the recent government changes, but they’re still really knowledgeable and they don’t know a lot about ocean, which. Is great, but the challenge you have then is if I am somebody who doesn’t know a lot about AI, I’ll go and hire someone who doesn’t know the domain and the likelihood of you finding someone who understands both and a very specific ocean domain that you’re in as well as artificial intelligence is slim to none. And so then, you know, you hire some younger person and you throw them in and say go. Do this or go fix this. It’s hard and I use the example often like back in the day. It used to be there, some would say we’re going to hire a data scientist to come in and fix our hour problems. Well, if you have an articulated what the HR problems are, they don’t understand your data. If you don’t have specific solutions you’re looking for. And then you know the fire the person after six months and say they didn’t do their job well, what are the challenges? And so the the biggest gaps that we do see it’s it’s it’s kind of matching talent where it needs to be. You don’t need an AI expert necessarily. It’s it’s people, it’s it’s companies trying to find someone who can come in and actually understand the strategy and and kind of build help and build that out. And that’s the part that isn’t always happening to your point, you’re trying to make payroll and have the right stuff. You’re not always. Thinking I might need a consultant to actually to come in and talk to me about questions that I haven’t thought of before, I go higher in those next people. 

Ben Garvey 

So that in itself is a huge skill set. It’s like it’s a very specific rare skill set that you and you know, a very small subset of other people could have, right? 

Ben Garvey 

It is. 

Jen Laplante 

Yeah. 

Ben Garvey 

How do you? I mean, there’s talk about talk about gaps. How do you tackle that? Are there tools that you use with groups to kind of pull the problem and the value proposition and the? You know, if you like the you know the the MVP out of out of the nebulous squirrel of possibilities that seemed to be available to businesses, do you, do you have a set sort of toolbox that you can apply to that and? 

Jen Laplante 

Yeah, it might last roll it deep sense. Where I was working with, you know, hundreds of companies on this, a lot of the times it was a bit of a round table. And so one of the things like so if your company and you’re looking to figure out what can I do, it really is problem identification. Can you first start with? What are the challenges that you have as an organization and what I used to do it kind of map out a bit of a grid and so your grids are going to have column to just like any kind of investment you’d Inc. Think about kind of, yeah, but it’s it’s like, OK, so I’m gonna. 

Ben Garvey 

Servaline canvas approach. 

Jen Laplante 

I’ve problem 1 so maybe problem one is inventory management and so the question is do I have data about this you know, could this actually save me money? If it could, how much money? So it’s almost like a rank order across various attributes. You know, could this how fast could it be if we actually did develop something that would help us better manage inventory? Is this quick for us to implement and use? Is there a lot of change management? But then maybe another pieces like, you know, maybe we have a really crappy scream and we’re not optimizing the data there. That might be a whole other opportunity for us as the. Money. And so I think companies to sit back and think about where they’re gonna see real return on effort and return on investment for these things because you know, you could go and invest in some AI tool that doesn’t really do a whole lot for you. And the the lean canvas idea. So once you know when I’ve sat down with companies and we actually found what those challenges or potential challenges they can overcome are, um, it’s called in a machine learning canvas and it’s kind of like, you know, that business model canvas, but specifically for where a machine learning project and the thing that’s awesome. Is it’ll set you know. It asks you kind of like your outcome, so one of the very first projects I worked on with a with a company was, you know, they were saying I want the model to be better than. The way we’ve ever done it. 

Ben Garvey 

Well. 

Jen Laplante 

A machine learns from data, so it’s never going to make a better decision than you did. But then it got into the question around accuracy. What’s good enough? And so it kind of set you out for if you’re going actually create some kind of an AI solution or adopt it, what’s a reasonable level of acceptable risk you’re comfortable with making? So maybe it’s 95% accurate and that’s good enough. So one of the projects we worked on was identifying the wave height in Halifax Harbor. So if the buoy at Chebucto Head went down. What would it actually be? So they understand. So if pilot captains are coming in, what when do? You make the call. Yeah. And so maybe 95% accurate. It’s good enough, or what’s your like? What’s your margin of error? You’re comfortable with and then the other thing that has happened, you know, companies need to think about are this machine learning. Canvas is good is what’s your outcome. And so if I said to you, I want you to predict the weather 3 days from now, there’s I’m going to, I’m going to decide if I wear a coat or not. Well, maybe what I’m thinking is I’m going to wear a coat or I’m not gonna wear coat. And so, because there’s only two binary solutions or answers or outcomes. It’s going to have a really the likelihood of having the right answer is really strong, but maybe you’re thinking oh, no, actually, Jen wants the numerical value down to a three decimal place and you go and you create this whole solution to decimal places and present it to me and like. But I just wanted to know if I was going to wear a coat or not. 

Ben Garvey 

Yes or no, yeah. 

Jen Laplante 

So some this this is like countless conversations I’ve had with companies around this where people are like, I want this. And so you have a business leader or a decision maker sitting articulating what they want, but they actually didn’t go down to levels. And so something like the machine learning canvas gets people sitting area. The table filling this out and saying what is the outcome we’re looking for, do we have enough data for this? Is it reliable data? Is it biased back to you know? Is there some risk that anyway, I highly recommend that as a tool of anyone’s ever looking at trying to map out machine learning. 

Speaker 

Yeah. 

Ben Garvey 

The. Canvas. 

Jen Laplante 

Yeah, exactly. 

Ben Garvey 

Louis Doran, right? And the kick. 

Jen Laplante 

Or an AI canvas. 

Ben Garvey 

Yeah. 

Jen Laplante 

Those are my absolute favorite tools. 

Ben Garvey 

Yeah. OK. 

Ben Garvey 

That’s great. That’s that’s. That was the tip the the secret tip I was. Looking for perfect. 

Ben Garvey 

Cool. 

Jen Laplante 

And you knew his name too. 

Ben Garvey 

He’s great. 

Ben Garvey 

I did. A little research, OK. Uh, so I mean, back to the to the talent piece. 

Ben Garvey 

What? What is missing and what would you? What would you kind of throw at a young person or not a young person? Somebody wanted to get into this. Wanted to get started or wanted to follow a path. Loves data wants to be effective in building. Models that are useful to the blue economy, let’s say. 

Jen Laplante 

There are a few things network I’m an introvert, so I understand what it’s like. It’s hard to get out and do stuff, so every time. 

Ben Garvey 

I find that hard. 

Jen Laplante 

Oh no, man, I’m going to be in my basement, hiding away from people after this. No, but like I I think one of the things I’ve done is gone and organized things. So you know, I organized a women in machine learning group and Halifax here. There’s over 500 members, so when people say there’s no women in tech, it drives me crazy because we have 500 members. Um, but there’s, you know. If go out and organize people, it’s all about the networking, and I think people forget that having conversations and getting exposed to new ideas and talking to people about what they. Do opens your mind to asking questions and learning more. And so if you’re a student who’s focused on data science or some kind of element around AI, they can really easily get hyper focused on like the Kaggle competitions as the website where you can get access to data and you can do lots of competitions for training the vast model. But that’s model isn’t. It’s not real world. You’re not going to get pure and simple problem. Pure and simple data. And so having students get exposed to what other people have lived in their lives when they’ve done those roles already is. The valuable I think you know. Volunteering in different groups, but I one of the other pieces too, is business acumen. And so I think the key differentiators that I found with students who are in that, you know, the computer science element has been those who’ve gone and taken classes on entrepreneurship or something around business strategy have really been valuable because I took a masters of computer science program at Saint Mary’s and years before I’d reached out to this masters in computer science program because you could have students do projects. And so I had a student do a project. Um and data when I was at Atlantic Lottery and essentially the recommendation that came back was you don’t sell enough products between 10:00 PM and 8:00 AM, so you need to do something to sell more then because it was essentially the student had just analyzed the time of sale. Well, that’s great, but no retail store is open between 10:00 PM and 8:00 AM. 

Jen Laplante 

We already had online gaming, like that’s not the solution, but because the student literally just went into data, analyze data for what they saw and just made what is not necessarily a reasonable recommendation. They came forward and I think it goes back to the. Piece that having a bit of understanding of how businesses work, what might be the priority and strategy for companies would be really valuable for them when they’re analyzing data and hopefully maybe helping try to find solutions or problem solve or flag something that might be a risk that the person who doesn’t necessarily know the data that well wouldn’t have realized. 

Ben Garvey 

Fascinating. And are there, you know, are there program, I mean you you said you’ve you’ve got this this group that you pull together, but are there other programs? That are generating that are bringing out those skill sets in young people here. 

Jen Laplante 

Yeah, I’ll give plugs for certain ones. There’s one sorry Hema women, so I’m going to harp on it. So there’s digital skills for women, plus Digital Scotia hosts that that is incredible for people who don’t have any tech skills really to help identify. But there’s some micro credentialing that gets offered through digital Nova Scotia. As well and you know that’s tough because it’s only for people in Nova Scotia, unfortunately. But there are meet up groups. I think the challenge with some meet up groups, you know, for example, Volta here in Halifax has one around generative AI, which is really good, but if you go in it’s very much amend thing. So I myself might always feel comfortable walking in because I feel like I might not know anything and I’ll be kind of shy. But there’s, you know, there’s so many great podcasts out there, too, that it isn’t always necessarily the event based piece, right? You know, there’s A at there’s the AI daily brief as a fantastic one. Today, a machine learning’s another one like there’s some really great folks out there. Another person I recommend people always follow as Gary Marcus. He’s fantastic. He’s very cynical, but very honest in terms of expectations of where AI fits. And so I think. Sometimes, too, it’s those different voices where you’re going to hear different perspective because it’s not deep on what is the latest model and how do you use it and how do you train it and where do you get data. It’s on the application of it and the challenges of it, and I think that’s the piece that students need to step back and think. 

Ben Garvey 

Read the. 

Jen Laplante 

Big and broad picture, yeah. 

Ben Garvey 

Bigger picture exactly and you haven’t plugged your own oceans AI. 

Jen Laplante 

OHP and I write, I write a newsletter. 

Ben Garvey 

Reply them. Newsletter. 

Jen Laplante 

I’m LinkedIn Ocean AI and I it’s funny. Thank you. Yes, anybody wants to subscribe, please do. I like sharing whatever is happening in the ocean space, but I try to always pull in something that’s happening bigger. Picture that’s not specifically ocean because I have a lot of, you know, friends and. Colleagues that don’t necessarily follow this space, and I think they need to be as meters sometimes around what’s actually happening. 

Ben Garvey 

I followed. It’s great, Louis. 

Jen Laplante 

Oh, good, thanks. Good. 

Ben Garvey 

Drink. 

Ben Garvey 

Drink my coffee. Read it’s good, good thing to do so. 

Jen Laplante 

Awesome. 

Ben Garvey 

I feel educated every week, so thank me for that. 

Speaker 

Good. 

Jen Laplante 

That’s good. Thanks. 

Ben Garvey 

Where’s it all going? Where do you think? Where do you think the ocean tech sector and the and the AI piece in particular is going? Is it a bright future? We got, we got lots going on obviously now, but does it is headed, there’s going to go. 

Jen Laplante 

It has to. It has to be right? I think that’s the only. That’s the only thing we’ve got going. I think the ocean space is really important. You know, we have an interesting four years ahead in the US and probably five years ahead in case. Yeah, I think that the thing that’s important in the ocean space is productivity. There’s a lot of jobs to be created. I think there’s a lot of work that’s going to happen when we think of ocean space. The environment is really important and our oceans getting warmer and our storms are getting worse and we need energy coming from the ocean. So I think emphasis needs to double down on what the ocean can do and as a result, you know when more money and investment comes into the space, more adoption of new technology will take hold. And I think that’s going to be the thing that pushes AI to the forefront because we’re not going to have enough people to do the jobs we need. And so we’re gonna have to find ways to automate and make things more efficient. 

Ben Garvey 

Yeah, I mean, that’s a great. That’s a great place to leave it. That’s fantastic. Really optimistic and I’m right there with you. These fantastic. 

Jen Laplante 

Awesome. Well, thanks for having me. 

Ben Garvey 

Thank you. Again, that’s a great conversation. 

Jen Laplante 

Thanks. 

Daniel Baugh 

This episode was produced by Me Daniel at Ingenuity Studios. If you have a story to tell about life within the marine sector or have an engineering challenge, you want to share with one of our experts, please reach out to info@enginuityinc.ca. 

That’s INFO at ENG-INU-ITY INC.CA and to learn how you can overcome your organization’s harsh environment challenges. Please visit our website, Enginuityinc.ca we’d love to hear from you until next time fair winds and following seas.