Charlotte Ward: 0:13
Hello and welcome to episode three hundred and twenty of the Customer Support Leaders Podcast. I'm Charlotte Ward. Today, welcome Craig Stoss, the anti-doom and gloom guy. Today I'd like to welcome back Craig Stoss. Hey Craig, lovely to see you again. How are you?
Craig Stoss: 0:39
I'm well, Charlotte. Great to chat great to chat again. I always love coming back on these conversations.
Charlotte Ward: 0:45
Yeah, it's so good to see you again. It's been a little while, but welcome back. For the benefit of new listeners, would you like to introduce yourself?
Craig Stoss: 0:52
Yeah, my name is Craig Stoss. I am the VP of Solutions at a company called Codif. We do AI agents across the e-commerce lifecycle, from shopping assistants through your support and policies and through to your returns automation and other automation activities. I've been in CX for the better part of 25 years in all sorts of roles from running support teams, consulting, leading children's computer camps, and now again very focused on the e-commerce customer experience.
Charlotte Ward: 1:26
That's awesome. That's awesome. And we are here to talk about a key component of your day-to-day work in some respects, aren't we? About AI and its application across the industry and everything, which has been a huge
Charlotte Ward: 1:40
amount of doom and gloom over the last year or so in support. But you're here to to pitch yourself, if it as it were, as the anti-doom and gloom guy, right?
Craig Stoss: 1:50
I mean, anti-doom and gloom is is I feel like there's a lot of doom and gloom in the messaging. You know, AI is going to take our jobs, AI's floppy is taking over the internet. You know, there's going to be also these problems with AI. And I I I just don't think a lot of that's true, or at least it's not going to be part of a mass adoption in the next two, three years. And and and I that's that's my take on the on the situation. I think I think that the I think that the media loves to scare people. I think that AI historically people are trained to think about the bad things about AI. Uh, recently just had a conversation last week about how people come into AI conversations, whether it be a bot or or a phone situation, and they're immediately looking for the negatives. They they they that they there's just been this training that people don't want to use AI and expect AI to be bad. And so they're pleasantly surprised when it's good. You know, my my view is that we need to make the AI solutions that we use day to day to a point where people expect them to be good better, expect them to be helpful and and not and not come in with that expectation that things are are going to be bad.
Charlotte Ward: 3:08
Yeah, I I'm inclined to agree with you on a lot of that. I do think the I do think we're I mean the media is definitely guilty of a bit of scaremongering. We're also guilty as a as a society of training people to have a certain perception, training compute consumers to have a certain perception about AI. Where do you
Charlotte Ward: 3:32
think all of that training and perception is most divergent from reality?
Craig Stoss: 3:40
Well, in order for a lot of this to happen, there has to be this this mass change in the way the world works, right? The way the way that AI is adopted. You know, think about the internet. You know, the internet kind of came out to at least the public view in the in the 992, 93 range, and and it didn't start really impacting mass amounts of lives until I mean, gosh, probably even post you know 9-11 in the in the North American context, right? I mean, it you know, we didn't have the the the websites and the news and the and the ad industries. I mean, this was 2010 and even later.
Charlotte Ward: 4:23
And
Craig Stoss: 4:23
so it means we're looking at uh a mass adoption was even later than that because you know it wasn't like you know the the the whole world was adopting this, it was you know, North America, the Western world seemed to be taking on, you know, faster and faster. And and now now, of course, it's a daily part of our lives. And I see the evolution of AI in the in the same way, right? Uh one of the examples that I use quite a bit, and I know this is a bit of a tangent from your question, but you know, I my my colleague has a one of those self-driving Teslas, and and they they use it in San Francisco. And I've been in a Winmo
Craig Stoss: 4:58
San Francisco, as I think you you and I have talked about um before. And and yeah, great technology, right?
Craig Stoss: 5:05
I I you know I I'm fascinated by it. I loved it. I I've watched this car move and how it made decisions, and I've studied you know how how it reads signs, and but yeah, that works really well in San Francisco. But I live here in Toronto, Canada, where snow cover covers the lines of the streets, it covers the signs. Self-driving cars are not coming to Canada in the next five, 10 years. Like there's not gonna be that mass adoption. So are we gonna see all taxi drivers lose their jobs, as the media might might say? No, we're not. Albo countries, you know, like India love India, but their infrastructure on their streets is is maybe not up to par with San Francisco.
Charlotte Ward: 5:48
Uh
Craig Stoss: 5:48
you could easily argue that. And and so India's not gonna adopt it, China's likely not to adopt it in the same way. And and so this mass adoption and the and the mass firing of all taxi drivers or all truck drivers that deliver our food is not going to happen, at least not in the in the foreseeable future, because the infrastructure doesn't support it. And so, you know, when you when I say when you ask about the the biggest divergence, to me that's it. It's it's that people are looking at these microcosms. And sure, AI can drive a car. I'm I'm not arguing that it can, and it should drive a car, it's going to be safer for us. I'm I'm convinced of that. But the mass adoption of this, you know, we're talking all the governments getting together and figuring out how to standardize the infrastructure of cars driving. And
Charlotte Ward: 6:41
I
Craig Stoss: 6:41
mean, that we're that's billions of dollars and and decades of work to make happen.
Charlotte Ward: 6:47
Yeah, and and socio-political change required to do that as well is huge, is huge. It would seem to be almost insurmountable if we're talking about global transformation at that kind of scale, uh, which is to your point required for global adoption of a single technology. Is that just a matter of time, though? Because, okay, the internet, I agree with your broad timelines, you know, mass adoption is is relatively recent from when you consider it as uh a trajectory from its birthplace. The same is true of most technologies if you look back over the years and decades and even centuries, I suppose you could argue, you know, the same is true. There are still pockets that haven't, you know, uh any technology you care to name, frankly. But there is a certain critical mass at some point in the life cycle of most technologies, I believe, where we could say, okay, this is broadly adopted to the point where it might be considered a kind of a basic human right. And I I think it's it's ludicrous at this point in the birth at the birth of AI to think of AI as that, right? But at what point did we start to think of electricity as that? And at what point will we think of the internet as that? We're already there with other like more fundamental, you know, the right to to water and to medical care and to, you know, reasonably efficient farming and things like that, are I think we'd have to say broadly globally adopted at this point. But but is
Charlotte Ward: 8:25
it just a matter of time, is my point.
Craig Stoss: 8:27
Yeah, uh certainly it is a matter of time, right? I think the time scales are different. But I also think what time allows us to do is exactly what happened with the internet. And I and I don't want to keep coming back to that parallel, but I think it's an interesting parallel. In that, what did I say, 20-year, 25 years time span that I just outlined of the internet? What happened? Technologies that relied on the internet came up. The meeting were we're recording this on Zoom. Zoom wouldn't have existed if it wasn't for the internet, right? And and that's true with jobs. You know, internet advertising became a thing. You know, there's a whole SEO. Now we have AEO. You know, these these are these are roles and positions that that were invented because the technology shifted over that 20-year period,
Charlotte Ward: 9:15
and
Craig Stoss: 9:16
AI will do the same. I don't know what my I have an eight-year-old son. I have no idea what he'll do as a job in in 10 years when he's when he's ready to enter the workforce or 12 years. But but the point is that the doom of gloom is saying, well, every job is going to disappear immediately because self-driving cars exist, or because AI is going to take all support roles, which, you know, more relevant to our world.
Charlotte Ward: 9:37
And
Craig Stoss: 9:38
I'm looking at it saying, by the time that happens, by the time AI gets good enough to take every support role off the table, something else, and I'm not the visionary that will tell you what that thing is,
Charlotte Ward: 9:50
but
Craig Stoss: 9:50
something else will come to backfill that that set of roles. And maybe there's fewer jobs. That's likely, but maybe that means that we have to start to your social socio-political comment. We have maybe we have to start thinking about compensating humans differently, like universal basic incomes. And do we pay hourly wages or is there a better way to compensate humans in an AI-driven world? Some suggestions I think Bill Gates is famous for this one saying, well, should the machines, you know, the whatever the machine produces, the AI produces, shouldn't the the value of that, the whatever the revenue from that go to the people that would have done that job somehow through some distribution system? I don't know. Again, that's not my area of expertise, but I think that that's that's my point is that this isn't going to happen in the next six months. This is gonna happen in the next maybe not 20 years, because technology is moving faster and faster and faster. So let's even say 10. 10 years is a long time as it comes to the evolution of jobs. I mean, even even thinking back to between when I started my my degree in computer science in in 2000, you know, to my kind of second, third job in in the in in now into the 2010s.
Charlotte Ward: 11:08
I mean,
Craig Stoss: 11:08
some of those job titles didn't exist before, you know, and and I worked with colleagues that were working with technologies that didn't exist from the point of me starting my my degree to the point of me being in the workforce. That's exactly what's going to happen here. And and that's why I just don't see this doom and gloom. And so, yeah, that that to me is the the biggest thing that people are forgetting is that you know, there's a there's the old, there's an adage I heard a while ago that you know there's fewer horses on the in the world now because we don't use, we don't breed horses to do all the manual labor that they used to do, but now we use horses for recreation, right? You know, riding is a big thing, it's a sport in the Olympics. You know, that that's the kind of and I don't want to apply humans to horses, but that's kind of the analogy, right? Is that we're going to have humans in the future. And how do we how do we use humans? I suppose is the way to just to phrase it.
Charlotte Ward: 12:07
Yeah, yeah. I I I think fundamentally, I I agree. Like I I think that there is some sort of whichever way you look at it, shift left, shift right, things shift fundamentally. There is space created by any of these technologies and the innovation they bring because they provide platforms or space on on which and in which new things can exist and and also be innovated, and along with them, job companion jobs and and opportunities for humans that we just can't see right now. I I do agree with you as well that I think that the pace is different, and I think this is why it feels particularly unsettling with AI, you know, when you look at the historic kind of you know development of some technologies from a century or two ago, you look at electricity, you look at some of the others, those were those felt seismic to the people living them, you know, but the pace of change was was slower, but the pace of life was slower, and the pace of adoption was slower. Technologies developed more slowly, were adopted more slowly, and we're evolved, you know, we're innovated on more slowly. And so what felt seismic might have taken 50 years to develop or longer. And we've seen that to your point accelerate with the internet. The internet we are looking at broadly 25 years of of like mass adoption of that bit of the life cycle that is that accelerated adoption anyway. And with AI, we're looking at a few years, right? We're looking at two, three, five, maybe. I think that's the discomfort, the pace.
Craig Stoss: 13:41
Yeah, and it's funny, you know, I uh one of the comments that I
Craig Stoss: 13:45
the one of the things that I observed, you know, I lived in in England for a while back in the in the 2010s, and I was in I was incredibly blown away by first chip and pin for your credit cards and debit cards that didn't exist in Canada when I had moved over. The the use of text in in England, text messages for paying for parking and and various bills that didn't exist in Canada when when I moved over. And I was just in New York City last week, and I'm still there are still restaurants using signing for your credit card, they don't have chicken tin machines. And and that's a that to me is a really good example, right? In the sense that I came to England in 2010 and there was this really cool technology. Canada adopted adopted it quite quickly as well. And there are still places in America that haven't adopted it. And this, and I'm sure that's true for many countries, right?
Charlotte Ward: 14:41
And
Craig Stoss: 14:41
and so even something as simple, a simple technology like that, where where it's literally just a different type of machine you need to use,
Charlotte Ward: 14:50
yeah.
Craig Stoss: 14:50
Isn't be isn't widely adopted. And that's you know, we're now we're now going on on another 15, 16, you know, years. And so there's also that element of it. And and I think with AI, there's another another thing that's interesting is is going back to kind of socio-political or moral implications.
Charlotte Ward: 15:09
Again,
Craig Stoss: 15:09
I'll use a car as an example because I think it's a really fascinating use case. If you're driving down the street today and uh and a child runs in front of your car and you swerve and you now crash your car into a into a pole, you know, and and you're injured or whatever,
Charlotte Ward: 15:25
no
Craig Stoss: 15:25
one would go and look at you and say, you made a a conscious decision to swerve that car. That's a that is a gut reaction, right? That is uh that is just a human, oh no, get out of the way. And and and so that would be how it's treated under law. In the AI-driven world, in the in the in the coding world, that's a deterministic decision. The the AI has to make a decision of do I hit the child or do I swerve the vehicle? And if
Craig Stoss: 15:50
the choice was to swerve the vehicle and therefore potentially injure or kill you, well, that's a big implication. Now who's at fault if that happens, right? Is is because it's no longer a gut reaction, it is a deterministic decision that someone made. And I'd argue there's a huge set of morals around that. Who gets to make that decision? Is it the government? Is it the car manufacturer? Is it Elon Musk?
Charlotte Ward: 16:14
Is
Craig Stoss: 16:15
it you? And if the decision is that the car could kill you, would you buy a car that in theory would would kill you? You know, would you buy that vehicle? And and so, and I think AI has a whole bunch of these moral things around it, right? And and so and some of it's you know, obviously exists today. There's self-driving cars today, and I don't know what how they're programmed in these situations. We're probably gonna find out at some point in the future what how these things are programmed, but the the implications of that are widespread, and and each country may have a different decision on who gets to make those software choices. And and just like you know, the EU has such stricter rules around privacy than than the than North
Charlotte Ward: 16:59
America,
Craig Stoss: 17:00
uh you know, and I'm not I don't have an opinion on whether that's good or bad. I think I think it's better for for society in many ways. But that's again a huge implication of of the AI adoption rate, right? And and so yeah, I think that the timeline is sped up, and I think that the you know we're seeing we're seeing some level of adoption, but there's still going to be roadblocks along the way as these things roll out and become, again, going back to mass mass adoption conversations. So yeah, very, very, I think that you're right that the speed feels daunting right now,
Charlotte Ward: 17:42
but
Craig Stoss: 17:42
I don't think that equates to overall fear of of a of a technology. We have to learn to adapt to it. We have to learn how to teach kids, children in in our education system how to ethically use AI for their work and how to how how to how to make sure they're not plagiarizing or that their fact facts are validated or or
Charlotte Ward: 18:07
you
Craig Stoss: 18:07
know they're not just skating by in a math, a math or science exam because AI wrote it for them.
Charlotte Ward: 18:14
Yeah, yeah. I I think the intersection of of morality with technology is is something that feels new here at this level, and part of that is the speed. I mean, there's definitely morality and technology questions around the internet, let's face it, and they are valid, but they are much more to your point deterministic. They are much more well, they are much better described, they are much better gated, and all of those things and legislated for in a way that we we're entering a kind of bit of a brave new world. I mean, that example you gave of where does the where does the blame lie for a car that crashes you into a lamppost opposed to a child is a really interesting one. And I I am one of the world's like, well, no, that's an overstatement. I am I'm gonna actually cut that bit out, as a huge overstatement. I am I am a huge advocate for the use and adoption of AI generally day to day. I just think there are no excuses to not engage with it day to day. However, there is a little bit of discomfort, even for me, around that kind of intersection of morality and commerce where, you know, the the car manufacturers kind of commercial decisions around that determinism, that that decision making, and how they encourage me as a consumer to buy a car that might kill me, I think is deeply uncomfortable. I don't know that I have the answer for how how I feel about that. And I I would be certain that there are car manufacturers out there who are thinking much more deeply about that than I have at this point, and are about to put a car on the road that already has made that decision for me. It's a very, very I can understand the discomfort, but I don't feel doom and gloom about it. I just feel like there's there's a lot of unknown, and maybe that's where the doom and gloom is for a lot of society right now, is in the is in the unknowns.
Craig Stoss: 20:17
There, you know, and I think the other thing that humans forget, and I think this is generally uh uh just a really fun human thing,
Craig Stoss: 20:25
is that the this stuff, this moral decision stuff, isn't related to AI in many ways. There's that famous story, and and I think it was referenced in in Charles Dugig's Power of Habit, but I will I also read about it in another book, and I can't think of its name, but it was written by a data scientist where Target, the the the American department store, was able to determine when women were pregnant before they even knew,
Charlotte Ward: 20:55
based
Craig Stoss: 20:55
on certain shopping habits. This was like, I think this is went back in 2010-ish. Like this was a this was a long, long time ago,
Charlotte Ward: 21:06
uh,
Craig Stoss: 21:06
in the in the scheme of of internet and AI
Charlotte Ward: 21:08
technology. But
Craig Stoss: 21:09
it was all just data analytics, right? And it was just they put a bunch of numbers into a spreadsheet and could figure that out. And and I if you haven't read that story, I would, I would absolutely read about that story because it's it's a fascinating study in in data analytics. Now, can AI do that faster? Yeah, and can it do it against more needles and haystacks than you know, is a woman pregnant or not? Yeah, absolutely. But the tech that the concept of these moral things, like should Target be monitoring, you know, your your you know, your your lifestyle, your your your your family planning choices and and
Charlotte Ward: 21:47
and then
Craig Stoss: 21:47
somehow capitalized on that? Is that a moral thing to do? And yeah, AI will absolutely make that more prolific. So I I I I love that discussion because it it's always existed. That the concept of marketing and and promotion and and companies trying to find that edge to sell something a little bit more that that's just the driving market of capitalism. And and so that's it's not new, it's just exacerbated by this technology.
Charlotte Ward: 22:15
Yeah, yeah, and and sped up. Absolutely. I couldn't agree more. I couldn't agree more. Craig, it didn't feel all doom and gloom. So thank you for that. I I think it is a I think we all have to recognize it's a fast-changing world and engage with it, learn about it, understand it, plan for it as best you can, but be prepared for change is my take on it. And I'm optimistic about that fundamentally. And I I get that change is scary for people, so that's where a big element of the doom and glue comes from, I think, in the industry and indeed in any any industry as they engage with this technology and this changing landscape. But but change change on average is for the better,
Charlotte Ward: 22:55
is my fundamental view of the of life.
Craig Stoss: 22:58
There will always be bad actors with any technology, with any with any you know, advantage they can gain. Yeah, you can't get rid of that. But I I am with you. I pa I I'm generally positive. I think this will improve a lot of things, you know, both in the world of science and medicine. I'm I'm super excited to see AI applied and more in the in the doctor's offices. I I want to make sure the security's there and the and the privacy is there. But overall, yeah, I have a very positive outlook of of this, and I don't see the apocalypse happening, you know, or or the the the the constant comparisons to Terminator and and uh uh and and us you know ending the world, at least not in the next 30 plus years.
Charlotte Ward: 23:47
Yeah, fingers crossed.
Craig Stoss: 23:49
Yeah, there we go.
Charlotte Ward: 23:53
I I I hope so too. I couldn't agree more. Thank you so much for for joining me today, Craig. It's lovely to have you back after a bit of a hiatus. And will you come back and join me again sooner?
Craig Stoss: 24:04
Absolutely. We'll we'll have to make it happen more than more than once a quarter.
Charlotte Ward: 24:09
Yes, yeah, absolutely. All right, thank you so much. I'll talk to you again soon.
Craig Stoss: 24:13
Thank you.
Charlotte Ward: 24:17
That's it for today. Go to customersportleaders.com forward slash three two zero for the show notes. And I'll see you next time.