Techstrong Gang – September 6, 2024

Happy Friday, everyone. Operations as code. Do we need another as code, the DevOps loop?

Is it really infinite? And is AI the future of e-commerce? We've got all that and more for you on Textron Gang.

Hi everyone. Happy Friday. It's Alan Shival for Textron Gang.

Welcome to our end of the week show, and what a show we have. Um, we've got some great topics. I teased it earlier, but not only we have great topics.

We have a full house of great Textron gang members. It might be one of our best lineups ever. Let me introduce you to our gang folks for today.

First of all, then they're all returning gang folks. If you watch Text on Gang, and if you don't watch Text on Gang, you could watch it on Textron tv. We have every episode up there as well as on the Textron tv, YouTube channel.

So please check that out. You could subscribe on YouTube as well and never miss an episode. Um, but let me get back to introducing my friend Lisa Martin.

She's the host of her old podcast at Lisa. What's the name of your podcast? Hey Ellen.

It's called Marketing, art and Science. And we really dissect with CMOs how they're pulling those levers and flexing those muscles of art and science and things like generative AI to really improve business outcomes. Absolutely.

I queued that up for you real well. Halles. Yeah, Thanks, Ellen.

Wing to you. Welcoming. Thanks for being here.

Um, Lisa, I should also mention is with the Futurum Group where, uh, does and analysis of marketing, uh, industry spotlights as well. Also joining us, well as part of FU Group and Visual Impact. He's their CTO.

He's also a longtime friend of us, of us here at, uh, tech Strong Guy. Courier Guy. Welcome.

How are you Doing? Great. It's good to be here.

It's good to have you back on Guy. Always good to have your input on our gang shows. Speaking of input on our gang shows, she's kind of our newest gang member and we are thrilled.

Tickle Pink to have her here. hopely DevOps, software development operational IT person. Yes, extraordinaire.

Hey, hope welcome. Thanks for coming. Thank you for having me back once again, Alan.

Uh, this is a highlight. I love the topics that we discuss and, uh, little rabbit holes. I get to go down when I'm Yeah, I know.

Look, it keeps it interesting, man. Yeah. Thinking of keeping it interesting.

Uh, he's the c our CTO here at, uh, Textron, but he's also Futur, CTA and he's been doing a bunch of writing and talking lately. And, and an original gang member, a founding gang member. Mitch Ashley.

Hey, Mitchell. Wow. I forgot to wear my colors.

Or maybe Black. I was just thinking that Maybe we should, I wanna put you on like some kind of, what do you call head bad data. Something.

Yeah. I need a, like, facial tattoo or something, don't I? Good.

Going on that. Okay. We're not, I'm trying to gag.

I won't propose any. Okay. Good To be here.

And, uh, love being here with, uh, Lisa Hope and Guy. Absolutely. So it's a great panel.

And then joining me here in our Boca Raton headquarters studios, unfortunately still lamenting yet another Yankees lost at the hands of the Texas Rangers. We were, this wasn't even a blown one. We were behind from the get-go.

Um, Mike Ard Sad Day Again. It's another sad day for you in Yankee Land. Right.

But I have a new title, apparently, according to Hope, since I'm put together some of these subjects. So that makes me master of the Rabbit Holes is that Master of the Rabbit holes. Maybe we should change your name to Peter anyway.

Wow. So it's Friday today and, uh, we're clearly we're allowed to take a little liberty like that, but let we, let's get serious. We've got some good stuff to talk about.

I think the first one is, um, we call it the Rise of operations as code. com and, and Mike, I'm, I'm, I want to throw it over to you to kick off. Uh, you and I spoke about this.

I was skeptical, but I think they won me over. You were skeptical. I'm surprised.

There You go. Shocking. Shocking.

Um, well, as far as I can tell, theoretically, we've been talking about it as code now for years, and we kind of take it for granted in a lot of spaces and when we talk about tests for code. But why not take it to its logical conclusion? Every operations we can think about can be codified and it can be managed as code.

And maybe in some regards, I wonder if we've been limiting our thinking. And then when I read this article, I was like, wow, you know, it kind of opens up, uh, a way of thinking that goes beyond DevOps or it takes DevOps thinking to people outside of DevOps and those processes. So, hope, I know you've been kind of floating around this whole DevOps category as long as we have, but yeah, what's your take on, you know, can we export DevOps as a philosophy to all kinds of operations?

I do believe that we can. I think, um, if you are truly speaking of the philosophy, uh, collaboration, rapid cycles of learning, all of those things, they are a benefit for anyone. But, um, if we're talking about how we are codifying everything, I think that does bring opportunities for organizations to work even more tightly together and have a better understanding of what they're trying to achieve and, and work toward outcomes.

Guy, let me ask you here as well, because, you know, we hear the term hyper automation a lot and everybody kind of nods their heads a little bit. And I'm not sure everybody knows what they are actually nodding their head to when they hear that term. But, um, you know, can we take code and kind of start automating processes in ways that we never imagined before?

And maybe we can do things in the age of AI that we would've considered impossible. And we're just short of imagination at the moment. I don't know if we're short of imagination.

Uh, this automate everything everywhere all the time. Idea. I was pretty old.

Um, you know, uh, my first thought is keep an eye on your test plants because now you are extending, um, you know, rapid development and CICD and all these wonderful speedy responsive things that you're doing with code, um, farther out from their landing spot, whether it's in the infrastructure, infrastructure is code now expanding up, uh, farther towards workflows and, uh, security SecOps, all that other sort of stuff. Um, but that means every single thing you do can have a potentially greater impact. The optimist to me, I'm very optimistic about, um, automation especially, uh, when it comes to, uh, uh, the, the increasing use of ai, generative AI in particular.

Um, I think actually, uh, cynicism and PEs pessimism, which is more my natural state, um, has told me for a while, nobody really understands this stuff yet. People are stepping forward without knowing what they're stepping into and all that other sort of thing. But I'm starting to feel that there's a lot of recognition of the dangers and there's a lot of hard work being done, um, to, uh, to automate and to insert AI more responsibly and ensure the human element, human review and that sort of thing.

So this, to me, it's like taking DevOps, like you say, in the DevOps paradigm, um, to this new level that's gonna actually be extremely helpful. Uh, and, um, I'll just come back to everybody. Remember, you still need to test, you still need to do qa.

You still need to be able to regress. You need to be able to do all of these things. None of this, um, makes any of that obsolete in the slightest.

No. So I, as I mentioned, I had a few thoughts on this one in Wasabi. Um, so That's from yesterday, just going back from yesterday.

com is just because you can't automate doesn't mean you should. Right? And I think that's a valuable lesson that we have.

Many of us in DevOps have learned the hard way, and they have the T-shirts and scars to prove it. Just because you can automate something doesn't mean you should. 'cause oftentimes guy, to your point, you automate, you go faster here, but it creates a warp or a bottleneck somewhere else because they can't handle that level or, or that, you know, velocity.

So speed can kill sometimes. Secondly, though, for those of you out there who are saying, oh, yes, just yet another as code, we have infrastructure as code, we have it as code, we have, this is code and that is code. What do we, what does this as code really mean?

Well, automation is part of it. Yes. Once you, once you boil something down to, uh, making it a code, it allows us to replicate that at a much faster speed and automated.

com, right? Puppet and Chef were, were the dominant, uh, DevOps tools because it allowed you to, to do, you know, spin up servers in AWS in essence, right? As code, it became a recipe cookbook.

And chef if, if you will. Um, so when we talk about operations as code, what are we talking about? Well, we're talking about taking that next piece of things, not necessarily a platform engineering.

And I do, I, I highly recommend you read this article because the author does a good job of saying, Hey, yes, platform engineering is part of operations, and it's setting up these guardrails that allow developers to go faster. But that's not really what we're talking about here in operations as code. We're talking about some of the more traditional operations after code is coded, if you will, maybe even after testing.

Is testing is done without, testing is never done. But you know, it, it's, it's a lot of that operations that I think falls a lot on SREs and stuff today, right? That's the new ops things.

And so the idea of reducing some of those tasks to code, which allows us to, you know, kind of have that gold standard to, to have that baseline that we can then rinse and repeat over and over and over and faster and faster and faster, it's not bs. It's real. And I think it's, I just never thought of it as operations as code till I read this article, and that's why I said I was skeptical.

But I think the article does a, a great job of laying out the logic behind it, how it fits into this whole arc. com in a long time if I say so myself. All Right, Mitch, how do we kinda put our arms around this speed issue?

And, and I'll give you a quote from somebody who I was talking to about AI, about this, and they just kind of looked at me and they said, you know, well, it's one thing to be wrong. It's another thing to be wrong at scale. So how do we kind of wrap some guardrails and controls around this thing so speed doesn't kill?

Yeah. Well, I think o one, one principle should we, should apply is the thinking that got you here is not the thinking that's gonna take you that to the next place, right? So I, I think what would would basically kinda give us lead boots, cement boots is to, to use a metaphor from one of our favorite Sopranos shows, um, is, is it's not about writing code.

We don't want operations people to go write a bunch of Python code or something, right? Um, this article talks about terraform a Terraform, you write code, you provide a configuration information that's written in something that looks like code, but it's just a configuration vial and then has an automation process to do that. And so, you know, doing things faster, to your point, if you really applied more than just automation, the principle behind, uh, a DevOps and lean and, and others is, think about it this way.

Let's, let's look at where the bottlenecks are. Where, where do we have things that slow down the ca to get done quickly enough? Or where do we have highly repetitive tasks that we're spinning valuable resources on?

Maybe we could automate and using some of those principles to identify where, where we can use, um, automation and, and operations as code. So, so kind of being smart about it. The other way, the way other way to think about it is it doesn't have to be code at all.

A lot of the operations tools now with the inclusion of, uh, early days of AI will identify, Hey, this is a workflow you're done in our tool before. Is that a possible automation? Is that a possible runbook we can create for you?

We see this pattern and, and it seems to be consistently on Thursdays that, you know, when the backups run, you go set these flags in this other configuration file, and then that informs some other process. Why don't I automate that for you? So it doesn't even have to be automation.

It can be, you know, no code version of automation as code, if you want to think of it that way. So we have to think about this, not just let's go write a bunch of scripts, um, to make our lives harder, but he but better maybe. So we're gonna have to jump to our next one, but before we do, you let the guy from Colorado do Sopranos things, whoever heard of Lead Overshoes, he meant cement overshoes.

I'm sorry, I, I'm outta my, I haven't watched an episode in a long time, but it's like, what is, like the 40 year anniversary of it or something? No, but Guy, I know you picked up on that right away from, from your roots, right? We, Well, you could start calling me Johnny Lips instead of Guy I, that gives you an idea where I'm coming from.

Pla River's only six feet deep, so it doesn't do much have led Over, Shane. I do, I do wanna mention, I know we're, we're, we're, you know, finishing up. I do wanna mention one thing, which is we have to keep in mind what the purpose of automating is.

I think the instinct is to say, automation let's me be more productive. Let's the team be more productive. We can do less in our imagin is we just kind of sit back and, like George Jetson, we tap the button all day long instead of, you know, having to work quite so hard.

But a really important element of automation is just consistency, so that those backups are done the same way every single time. Um, that does not absolve us from the responsibility of supervision and attention. But, you know, that's, to me, Alan, sort of one of the limiting factors in terms of whether you should or not, is understanding that benefit.

Absolutely. All right, let's take a break at this point. We're watching here, text Gang, and we're gonna come back and talk about that.

You know, we've all gotten used to that DevOps infinity loop, right? That we've seen a million times on slides. Mitchell says, don't get caught in the loop.

You're watching techron Gang, Discover Techron Group, the epicenter of tech innovation. We are your go-to for reaching IT leaders and practitioners worldwide. Our secret impactful content that sparks awareness, engagement, and top quality leads with us.

You'll access editorial websites, streaming videos, virtual events, custom content analyst research, and more. Join our satisfied clients. Let's revolutionize your tech journey.

Contact us today and tell your story to the world in the most powerful way with Textron Group. Hey, everybody, you know, sometimes there are things that are around that we just take for granted, and everybody sees it every day and they just nod their head and they don't think too much about it. And well, this DevOps loop thing has become one of those things.

We all see it, we all know what it's supposed to mean, and we all nod our heads. com that will then take you to a link of the video of Mitch having this discussion. But Mitch, um, you're kind of saying this loop thing maybe is, uh, shall we say, not so infinite and maybe messier than people think.

It's not actually a loop. So, so lemme tell you why I, this is one of those things that, like, this has always really bugged me. Like, you know, the Mike Ard, this has been bugging me, right?

And I'm looking, I've looked at that infinity loop for a long time and thought, you know, it, it, that isn't how software's created. That loop is just a linear task, a linear, linear series of phases. And I, if I took a tape measure or a yardstick and folded it into a circle or fold it in a circle and then flipped it to get an infinity loop, it's still a yardstick, right?

It's still that process. It makes it look like it happens in that order. And, and it really doesn't when, because you are doing smaller, when the principles of, uh, DevOps is doing smaller pieces of work, we can do that, especially if we're doing cloud native, we're doing smaller pieces of work.

Um, being able to do the, in a distributed fashion, whether it's amongst people or geographically, whatever, is inside each one of those steps, whether it's design, build, uh, test, et cetera, all the way down to deploy and operate is, there's a lot of little tiny loops that are happening in each one of those steps all the time. So, 15 developers are sitting here writing code and checking in and running an automated, uh, continuous integration when they check that in. And that's happened maybe hundred times, maybe thousands of times a day.

Um, and each one of those is their own little loop of doing some development, doing some testing, checking it in, blah, blah, blah. At some point that moves into the next phase, and that, that those pieces get tested again and some automation, those get combined. Maybe they get released together into production, et cetera.

So when you really look at how software is developed in today's modern approaches is it's not a linear process. It's a lot of executions of steps within that process kind of flowing down to an endpoint where it makes itself into, uh, production into some environment that it's then operated in and it's continually updated, right? You may updated it's being pushed out and AI models and new fixes to code or vulnerabilities.

So that, that's why I've given this talk a couple times is, and the infinity loop is nice. It, it kind of differentiates DevOps, but if you really think about how it works, don't teach your team just to do a linear, um, a linear process faster. It's about doing smaller steps of work, uh, being able to do those, um, independently, and then combining where you can collaborate and combining them where you integrate and test and do all those steps.

So, Hey, hope I, I've been waiting all day for this, but, um, is Mitchell in loopy on this or what It, you know, it, it brings to mind, uh, a training I did once, uh, at an organization for an Agile transformation, and there were a variety of people in the room and on the screen I showed the infinity loop, I showed typical waterfall cycle. I was talking through it, you know, by the book. And a person in the audience in the class raised their hand.

They said, but isn't it actually the same thing? I said, you know, it actually is. It's just in shorter increments.

That's, that's, that's really, that's really it. You, you are effectively repeating sort of the same cycles, but the the key is in, in shorter increments. But one of the other things that came to mind, um, with the DevOps loop is I have seen teams like, like those teams I was training, uh, who are going in and they are implementing DevOps practices, but people are saying, what else has changed in the organization?

What benefit are we getting from this? They say they're doing something different, but nothing, nothing actually looks different on the outside. And I think some of that is the disconnect from DevOps culture and the over rotation to saying that DevOps is just about tools and automation.

So, you know, look, listening to this conversation, I'm reminded of a, something I read once that if you took a, a strand of DNA, like a chromosome, right? And you in, you know, they're all coral Dre. And if you laid out a chromosome in a straight line, to your point Mitchell, about making that straight lining into a loop, it would be long enough to go around the earth, one little de one little chromosome in your, in one of the trillions of cells in your body.

We took that DNA and laid it out, you know, with chromosome laid it out. It would be like enough to go around the earth. So you could pack a lot into a loop, right?

But I think that is the beauty of DevOps, and it's not just DevOps. 'cause keep in mind, DevOps takes up where agile leaves off. And when we talk about doing things in bite-sized chunks, right?

It's agile, 25 years old, that is kind of brought that to development, right? The sprints and all of that. Um, and so, you know, it's the old story of how do you eat an elephant one spoon at a time, one spoonful at a time.

And, and I think that's the beauty of agile and DevOps and looking it at it in a loop that yes, maybe it does have all the things that we have in that waterfall, but the concept of doing them in bite-sized chunks and 1, 1 1 spoon at a time is, is the, is kind of the special sauce in, in my mind. Anyway, Lisa, I'd love to get an outside opinion here because you've been on this show a few times and you've heard us talk about DevOps and um, I guess, you know, we try to explain this and it's not pervasively adopting. Yeah.

And everywhere it's deeply in every enterprise somewhere, but it's not like the dominant way everything is done. From your perspective, does DevOps have something like, oh, is there a marketing problem for DevOps? Or we, or we just not having the right terminologies and the things to explain this to people that get more oomph going?

Communication is absolutely critical here. I think it's a marketing opportunity. Marketing can be the lingua franca between, uh, the development teams, the operations teams.

I think marketing can be a facilitator of DevOps, evangelism. Where is it working? What are those fight science chunks in which we're seeing successes and talking about the, and reinforcing the organization's commitment to DevOps practices that are successful.

That can be content creation and blogs, videos, training materials, et cetera. I think another thing that marketing, where marketing can really turn this into an opportunity is engaging with, uh, internal influencers where they are having successes and what does it mean to the business? What does it also mean to the end user customer?

So that storytelling piece, I think is where marketing can come on as that lingo and help the teams really clarify what are we doing here? Why are we doing it? To the points being made earlier, what are the benefits that we are seeing as a company?

How does it benefit, uh, our stakeholders, um, our customers, as I mentioned investors, the board, what's the value to the business here? And to your point, Mike, I think it has to be crystal clear, and I think in some cases, um, maybe, uh, what I'm proposing is cart before the horse. But I do think there's an opportunity from a marketing perspective to really start, uh, turning the knob on this into terms of being more widely adopted with more success stories, validating it.

I hope you've been, uh, you know, on the sales side of this DevOps motion, what is the challenges with the whole marketing motion as, uh, from your perspective and given your history? Um, there seems to be plenty of instances of use cases and plenty of examples, but I think the storylines are always kinda like, you know, we, uh, released more code faster and everybody kind of looks up and goes, so what? Yeah, I, I think, I think that that storyline is often driven by people in the organization who believe that there is a productivity angle to it that relates to speed and velocity, right?

But what they should be looking at is, are we building the right things that the customers want, they see value in and they are actually using, even if that means we, we kill, uh, you know, some of our pets along the way because the customer sees no value in them. So why should we build it? I think so many organizations, as long as we've been talking about DevOps, as long as we've been talking about the foundational pieces of DevOps that are rooted in agile, the point is still missed.

That you are doing all of this for someone, whether an internal customer or an external customer, and they are the arbiter of value. So, uh, shipping more trash faster is, is not an achievement. Shipping more of the things that they see value in is actually the achievement.

You know, one of the things I think to that point, hope is what, what's a little bit deceiving is, you know, it's why is why DevOps is often pegged as a tool or technology when it's really not. It's about how you create software. Mm-Hmm.

And, you know, I've done three DevOps implementations in organizations, and the consistent theme is the first thing you have to kind of address and work on, because this is about learning. It's about te learning together, how we're gonna do things a little bit differently. And you have to, to get past some hurdles.

And one of them is, um, fear of failure and frankly, um, when there's been situations when there's a penalty for failure and also a, um, striving for perfection or too much perfection. 'cause there's a, you know, the thought of mind of, I'm not gonna release this until it's fully ready, right? I don't want anything going out with any bugs or anything, and it's gonna take us this long and that's why we take nine months to release a piece of code.

And when you tell people that, no, what we're gonna do is we're gonna take that, we're gonna incrementally release this with pieces of functionality, and over time we'll have all that functionality in it, but it'll be through a whole whole bunch of successive smaller releases. And that way we can do those in much shorter cycles. That's, that's a incongruent thought when you're used to the, when you're used to the no, it's gotta be right when it goes out the door and it better be damn right.

You know, or else my job's on the line. So there's some real hurdles of learning you have to get over. Mm-Hmm.

Um, and not everyone can do that. Not every, not every organization's as ready to do that. And so you have to take where the organization is and say, how do we not, how do we apply DevOps principles?

Maybe that's still, you know, we're not ready to get to tackle that problem quite yet. So let's work on this. Take the bottleneck idea.

Let's look for where are the biggest bottlenecks and how can we improve flow of work and things like that. So, But your point of releasing incrementally in bite-sized pieces, right? The goal there is that customer, whoever they are, if you have built 50% of what you've planned and they say, you know what?

I'm happy. That's where you stop. You don't say, oh, but what about the other 50%?

We want to build it, it has no value. Now you can rotate and, and do something else, but so many organizations, um, uh, you know, maybe it's part of maintaining the silos, uh, that some people feel tied to, but it is really, really hard for a lot of organizations, honestly, to put the customer first. There is a lot of internal interest, um, that unfortunately blocks that.

Well, and you're working against that. Here's all the stuff we want, whether it's a product manager, what a customer might want. Yeah.

You know, but if you pose the question to 'em, if I could get you 20% of that in four weeks Yes. And then give you the rest over time, as opposed to waiting six months to get it off. Yes.

Would that be, would you be okay with that? Nine times outta 10, they're gonna say, hell yes, ac. Absolutely.

Exactly. That's a great one. And they Might, How could we do that?

Right? They're happy. So That's that change of frame of reference of how thinking about it differently.

Very much Fair. I think we're ready to move on to our next topic. All Right.

We're watching TechOne. Gary, we're gonna take a break here. We'll be back with the future of e-Commerce, but I won't be, I, I need to run out to a webinar, so I'm gonna hand it over to very, very capable Mike Azar Double.

Have a great weekend, everyone. All right folks. And we're back for our final segment.

And we're talking about e-commerce and the impact AI will have. Um, there's an article up on Textron AI about this. We invite you to check that out, but I'm gonna go to Lisa first on this.

'cause I have to say, all these e-commerce sites I see out there, they all seem to be the same. They're kind of boring. And I can't help but feel that we're in some sort of rut here.

So can AI do something for us here? And what might it be? Lisa, start with you.

Definitely. It's, I don't feel that we're in a rut. Um, I feel that the pandemic was this, I always call, call it a catalyst.

Ellen was talking about DNA in the last segment, and I'm like, did he know why he studied molecular biology and chemistry? So the catalyst comes, yeah, comes right to mind. I think what we, what we saw during the pandemic was this catalyst to really turbocharge this, um, online, uh, e-commerce experience.

And while, um, things came back to normal, the appetite that customers had for this experience being really customized, really personal, real time relevant is only going up. So what organizations need to do, if you're feeling, or if their customers are feeling, Hey, we're in ru, these sites are all the same, they need to really work on differentiation. Uh, the article mentioned a couple of great examples of retailers who are doing just that.

And I have to say, I completely agree with the Sephora reference. I'm a Sephora fan. I got a whole bunch of it on here today.

But they have a great experience where they're really facilitating the omnichannel experience. And that's what organizations have to do to differentiate. I wanna be able to engage with the brand, and Jenny, I can help feel this, whether I am in the store, on the app at home, and I want the experience to follow me where I go channel by channel, so that ultimately I'm getting not just the products and services that I want, but I'm getting these recommendations because they've leveraged gen AI to do customer segmentation analysis.

They've leveraged it to do sentiment analysis, return prediction, so they understand much more about how the individual is interacting with them to deliver a much more fine tuned and differentiated experience. And I think that's the opportunity that organizations have. Guy, what's your sense of what's going on here?

'cause I know you've worked on a lot of marketing stuff over the years, but, um, what's your assessment in the state of e-commerce? Well, I think Lisa's take on, whether we're in a rut or not, is when I would, you know, I'd follow her on that. Um, I think that the, the potential for this is maybe bigger than in almost any other industry, because generative AI in particular, it's kind of like a, I don't know, a communications amplifier if used right?

And if used right is obviously a really important part of that phrase, communications amplifier. Uh, meaning, uh, very much what Lisa was talking about, about an ability to provide more customized experiences. Um, because generative AI is not just using the one model that's trained on the internet and just regurgitating the same stuff.

There's all ways that you tune and use RAG and so forth to, to add a lot of context. And when you're adding, uh, contextual information about the, the session and the, uh, the, the shopper or customer and all that other sort of stuff, now you're really getting to this point where a person can interact in a very focused way and get that boutique feel from e-commerce. We are not anywhere near there yet, but it's clear, um, what kind of, uh, uh, potential generative AI would have to, uh, listen on the last segment.

We, we talked about, don't, don't forget the whole point here is that there are customers and users for these automated systems. Um, that's the end goal. Not to just throw them a whole lot of stuff, um, at great pace and at great speed.

So if we take the same principle here, um, I could see in two, three years really rewarding and satisfying shopping experiences that are also rewarding and satisfying marketing experiences for the retailers themselves to, to help speed up commerce and make it more productive and make it more useful for everybody. If that's happening now, um, it's only happening in bits, but it's definitely the potential and we will hit some bumps along the way. Hope, um, both Guy and Lisa are suggesting that I might be suffering from attention deficit disorder here, and that there are more things going on than, uh, possibly I'm appreciating.

But, um, one of the things I do see is, you know, there's always bots on these systems, and they're all kind of maybe first generation, but they're not very, um, shall we say, rewarding to engage with. So as we look at the future of ai, I mean, do you have an expectation or something? You're kind of looking at these sites and the bar's gonna get raised in your mind because if they don't have some sort of AI capability, you're gonna be, well, this is, you know, somebody's site from, you know, last century.

I think. I think that's accurate. And one example that immediately comes to mind is if you're on an, on a site and there is a box for you to get help or to chat with someone, the worst thing that I've experienced is for it to come up and say, we only staff this between these hours.

And if you leave a message, someone will get back to you or, or, or contact us again within our business day. Well, you know what? My business day, in their business day may not align, but if they have refined a really good chat bot that is staying up to date based on the types of questions that customers ask, it knows the inventory, it understands, uh, shipping policies, return policies, all of that.

Now I can get my answer. I, I'm a happy camper and I'm moving on and maybe, you know, I am an outlier. I don't care if it's a real person or not.

I just want to get my task accomplished. And extrapolating that a little into the real world, I was on a road trip a few weeks ago and I was so hungry, and the only place I saw to stop was a Bojangles. Well, let me tell you.

I pull up and a voice asks me about my order, and I realize, oh, this is not a human, it's the ai, but let me tell you, the order was so fast, it was so accurate, made me think maybe I should go to Bojangles again. I don't know. I'm not that much of a fan, but it was just so seamless and professional, right?

So I think a lot of it is going to be, what is my experience with this? Do I still feel that I'm valued? Do I still feel that I've gotten the right focus bingo.

And am I getting what I need out of the interaction? If they can hit that, yeah, it's, it's, it's a gold mine value. Yes.

All right, well follow that through for a minute. Um, are we getting to the point where soon, um, we may prefer the engagement with the digital human versus the normal humans that we might have encountered in the e-commerce support services because, um, that AI agent will be more precise. I mean, is that kind of where the next thing's gonna be?

I mean, what is the gonna be the role of digital humans in this whole e-commerce function? I don't know, Mitch, you wanna jump in here for a sec? I, I think there are times when we change modalities.

You know, we go from in-person shopping in store to a digital store. Well, having a digital representation of a store hasn't really panned out as a great interface, great experience for customers because it's not, doesn't have the same fidelity. Um, but there are other ways to do it.

I, I think the rut, I'm gonna take the rut position, Mike, I'll back you up. I'll, I'll give you some support, buddy. Um, the rut we're in is we're, we're tired of being inundated by ads.

We're tired of being, getting emails, and we're tired of going on a website and we don't click on banners or we don't, whatever, all that stuff in the side, you know what, I don't really care about all that stuff. I'm here to get what I'm looking for. What I think is smart.

And it goes to the, the customer experience. Just, you know, the pandemic was the catalyst to accelerate a lot of things. But to our earlier point, it could have accelerated in a really bad way or it could accelerated in a really good way.

And part of the focus needed to be, and still needs to be on digital experiences, digital customer experiences, not the customer's digital, but the experience is. So, what I mean by that, a good example of this, by the way, I don't need a 53rd bot. I already got so many dang AI bots.

I mean, every application has one. And I can't figure out what the difference in how to use each one. 'cause they all operate a little differently.

So we got, we gotta fix the bot problem, bot problem. But the other is, you know, let people consume information in the way they like to consume it. One of the ways I I've really grown to like it is I don't like ads thrown into Facebook or my Instagram feed, but I love articles about 10 top 10 travel tips for the holidays coming up then.

And read an article of someone who's curated content to say, here's a great thing for packing your bags and, and packing your clothes in your backs is another thing for protecting your, your, uh, RFID protection of your, your, um, your visa card or whatever it might be. Those I like to look at because it's like, that's a topic I'm interested in, someone's curated. Now, where AI think I think can come into this, and we're, we're clearly not here yet, is with a context graph, AI can map you to that information and curate it for you rather than a human always having to curate it.

So, for example, if, you know, I'm, I've been on Amazon and I bought rain gear for motorcycles in June, suddenly outta nowhere, I'm buying motorcycle gear. Probably means Mitch likes motorcycle gear. He might be interested.

Matter of fact, based on what he's buying, he's buying touring kind of gear. So he probably has a larger motorcycle or some way he's, he's riding a bike now. Um, we noticed that he moved from Seattle to, uh, Denver, so maybe he needs some cold weather gear.

It's coming up on August and September. Okay, well, based on what he is bought before this might be of interest. You know, it, it's, it's more than just what I bought before.

It's the context of maybe where I live or the time of the year or what I've been reading about. You don't want to get into the creepy factor, right? Yeah, that's, uh, I was, you Know, there's a, there's a balance.

Yeah, you gotta, and maybe, Oh, I know you just moved to Denver. What? Well, You noticed that, See, there's a level of trust with that.

So if you build trust in that experience, okay, I'm, I'm willing to, to let you know, be comfortable that, you know, that I moved to Denver. If all of a sudden I'm throwing stuff up of here's now Denver restaurants, it's like, okay, this is not helpful. I didn't ask for Denver restaurants and I just moved here.

And why is this thing given that, that's what I mean is throwing stuff at us, but knowing why, because it's interacted with me and it's helped me in other ways. I'm gonna help you in similar way. Now here, it's gotta build trust.

I'm completely with you, Mitch, but it's funny that you, uh, that you said you're gonna take the side, you're take the rut side and your, your finishing is like optimistic. Here's how it could work. And in, in your discussion of it, I think you pushed me onto Team Rutt, but I think I'm on team Ru, you know, uh, uh, going forward, I think that that what happens here is, um, a lot of very interesting activity that's not necessarily productive, but, you know, then we hit the ru we're gonna hit the ru in six months, nine months, whenever it is.

Because these, uh, AI services injected into the shopping experience are gonna start to, to annoy people. Because we all have to learn by doing, including the people developing these things. It's gonna annoy, it's going to concern, it's going to get headlines.

That'll, that's the rut that's coming. So I don't know if we're in a rut right now, but I have a different vision of this e-commerce thing, and I think it's gonna play out. I don't want to use your bot.

I don't care what bot you put on your website. I wanna use my bot. The bot that I'm gonna have is my personal bot that I'm just gonna go say, Hey, I need a pair of jeans.

Go find them and tell me where they are and have 'em delivered tomorrow and charge the card. And that is the end of my e-commerce shopping experience in my mind. Because, uh, and maybe it's just, maybe I'm a guy and, but there's a lot of folks who are like me.

It's like I go to the mall, I'm going there to get jeans, pull up car walk to where the jeans are, pick up the jeans, go to the cashier leave. So there are folks who want to have that whole shopping experience and it's a recreational activity. But I'm just saying that the current website experience isn't compelling enough for me to give up my personal bot to go hang out on your website.

So, And why do you want that personal bot? It Because it understands you. You trust it, you know it's not gonna take advantage of you.

Um, a little, little different in your case. 'cause if I remember, you only shop on Yankees ticket sites and Yankees, merch, so you're a little bit of a narrower mer use case edge case, I guess we should say. But teasing you a bit.

But that, that, that idea, you know, as one potential future, having that personal thing is because that personal is, 'cause you've got, you wanna say relationship with it, but you've got trust in it, you understand it, it's learned enough about what you want to be effective at helping you as opposed, opposed to the 53rd bot That's annoying on this site too. And now, so Now open me, nailed it with the trust piece and the value piece from a customer's perspective, that those have to be there a hundred percent for it to be successful and to evolve, Right? So now you're, you're saying, I've got a digital twin AI agent who understands my, my values.

Who knows, I've already purchased a pair of swim fins, so don't show me 50 more pairs of swim fins in hopes that I'm gonna purchase more because I'm not. Um, let me tell you, whoever, whoever gives me that little perfected AI bot, uh, AI agent, digital twin that can sift through all of the things, um, that, that's beautiful because friends ask me sometimes to go and hunt things down for them because sifting through all the information, but also if you think, um, my mind is going back to cosmetics, there's, uh, the environmental working group, right? They rate cosmetics, beauty products, all of these things, instead of having to go to their site, which would be a bad thing for them, I'll admit, but just saying, instead of needing to go to their site, now, my little AI agent can say, oh, you're looking for mascara, but you want it to be highly rated on the environmental working group.

You have all these other factors. Found it, bought it, it'll be at your house in two days. Bravo, bravo.

Here's where my little dream falls apart in my mind. Um, so I'm going to have a bot that I'm going to optimize to find whatever it is I'm looking for at the lowest price I can get. The other people are gonna have bots that are designed to sell this thing at the highest margin that they can get.

And our two bots are going to meet somewhere in the ethernet and battle it out for control of this thing and then ground each other to a halt because they're both optimized for the exact opposite thing. And what they, what are they gonna do then? Send me an email saying, can you sort this out When you send your bot to fight training camp so it can be a little more effective at fighting the other bot.

I feel like it, it's like the, uh, it's like the trading floor, right? Fight, fight club. It's like the, like on nasdaq, right?

Selling and buying. Good point. It, it's, it, it would become something similar.

We need bot Fight Club. That's what we need. Oh, I I, I like that.

I'd watch that show. I'd go to a website to watch that. You have robotic versions Of that.

Well, if it's happening in the ethernet or internet or Ether or wherever it's happening, uh, we'll need a visualization tool in order to see the blood gore. There you go. All right, Lisa, last word on this.

Are we, are we daydreaming here or do you think this is how this might all play out? I think that, and I don't think there's daydreamy going on. There's a lot of different perspectives here on, on this particular, um, episode.

I look at it through a, uh, an optimistic lens. Uh, I wouldn't say rose color glasses, but I think I, I really liked what I talked about the omnichannel customer experience and how we have this expectation as consumers that it's going to be that in a non creepy way to Mitch's point. Um, but it need, we need to, what companies need to do to differentiate themselves, as I talked about earlier, is they need to earn the trust that Mitch was talking about.

To hope's point, there has to be value in it for the, the consumer. There has to be value in it for the business. They have to be able to demonstrate ROI and value.

And I think there's a lot of opportunity there. I really do, I think. But I think we're at, in early innings.

I think we're at the, the kind of, the tip of the iceberg. All right, folks. Well, I'm not gonna get a hat, but I would like to see maybe let's make shopping fun again.

There you go. Um, I want to thank all our guests for sharing their knowledge and their insights. I want to thank you all for spending time with us and watching the show as always.

And we have a great lineup of content coming up behind us. And of course, you can find this content on Techstrong tv, the website we invite you to go check that out. Until then, we'll see you next time.