Key strategies for integrating AI in business – a FLOW Summit replay

TRANSCRIPT

Introduction to AI impact on business processes 

Jared Shaner (00:08):

Thank you. I know we started lunch a little bit late, so I appreciate you guys coming out to learn a little bit from myself as well as two of the most smart gentlemen I’ve had the opportunity to meet in the recent year. So framing our conversation today, I think both myself, I work on the agency side for an agency called Trellis. Don’t want to get up on my soapbox too much. Happy to discuss it later, but as I’ve looked out at the industry over the last year, and I’m sure a lot of the merchant retailers in the room and throughout this conference have felt it as well. We’re all seeing the effects of the macroeconomic changes, the changes in spending behaviors, and the idea of trying to do more with less in conjunction with this. Luckily, and I’ve been trying to keep up with the ferocious pace.

(01:01):

We’ve seen technology and specifically ai, whether it’s that hot button chat, GPT or wherever it might be, equally drive as quickly as a pace. And so some retailers are getting ahead of the curve. Two of them I’ll be inviting to the stage here in order to leverage AI and automation. I know for me it’s a little bit like a foreign concept. I’m still trying to keep up with it, but they’re going to share some stories today about how they’ve leveraged these tools to drive better efficiencies, really empower their teams, both at Abercrombie as well as Adorama. But without further ado, I’d love to invite to the stage my two new friends, Antonio from Abercrombie and Fitch, as well as Hani from Adorama. Excited to have this conversation today. I don’t want to do you guys in justice, so I’d love to have an opportunity if you don’t mind sharing a little bit about your background as well as your roles within your respective organizations.

Panelist introductions

Antonio Colicchio (02:05):

Get the honor of having the mic today, some technical issues. So hopefully this works. Well just shout if you can’t hear me out there. First of all, shout out to my fellow elevator survivors this morning. Anybody else stuck on the elevator with me this morning? Raise your hand. I think we were about five minutes away from deciding who we were going to eat first, so I’m glad the doors opened up when they did. That was a special moment. Sorry. So introductions. Yeah, so Antonio Kalichi, vice President of customer care and Automation. I’m also responsible for digital fraud for Abercrombie and Fitch. I’ve been with the business seven years, formerly Victoria’s Secret, formerly DSW. I’ve been in the space my entire career and really excited and looking forward to talking to you more about the journey that we’ve had around automation for us, specifically robotic process automation. So that’s what I’ll be focusing on during the session today.

Hani Batla (03:02):

Wow. Hello everybody. I can’t be as good at standup as Antonio here, but yeah, hi, I am Hani Batla. I’m the C-I-O-C-T of adAdoramaiam. I’ve been with the company four years. I have the pleasure of us introducing Adiam because I feel it’s the biggest small company most people never heard of except for a lot of the folks here. I see the professional photographers here and the videographers back there and they’ll all know who we are and that’s a good thing, right? Because a Rama at its core is a company that’s about catering to creators and that’s everybody as far as we’re concerned, right? It’s the professionals here in the room, it’s the folks at home, it’s the folks in the office, it’s you guys. You’re all creators of something. But AAM is a unique company because not only does, it’s one of the premier destinations for all things professional audio, video and photography.

(03:50):

We’re a fairly diverse business. We have other brands, scuba.com, number one, destination for scuba diving, sunny sports.com, all things outdoor camping, hiking, lifestyle, a professional printing business under boutique, a rental business under Arc cinema enabling your next Netflix series or mid-budget movie that I may come to a theater near you. In my role, my team is responsible pretty much managing all of our e-commerce experiences. We are 95% and enabling all of the experiences we deliver to our customers. And those customers I’d qualify as two key personas, creators on one end and adventures on the other end between the two of them.

Jared Shaner (04:32):

Absolutely. We had the opportunity to actually meet up last night and quick plug, if you haven’t checked out, M had a 60th floor restaurant absolutely amazing. And with the beautiful backdrop. Two things that I learned about these two gentlemen is that their heart, they share a lot with me on the agency side, which is they’re tinkerers at heart. We’re all sort of children, whether it’s professional photography or I grew up around Abercrombie and Fitch. And so it was great to see how you guys are constantly trying to innovate. I’d love to hear, because it seems like you guys are on the front lines, when did you make that decision to start looking into automation ai? And for those who are considering it, what are some of the telltale signs that maybe it’s time to start thinking about it?

Antonio Colicchio (05:21):

Yeah,

The ‘why’ behind automation and AI 

Jared Shaner (05:21):

Happy Dunno, if you want to go

Antonio Colicchio (05:23):

Shout to why. I mean the why is always is one of the most important things. The gentleman who was speaking about strategy was alluding to that, and I don’t think he overtly said focus on the why, but that’s what he was touching on. And so I think for us there’s many whys, but the big why that stands out is the why that’s always been there. And now it’s increasingly important in an inflationary cycle. And what I mean by that is the why of just net productivity gains, net positive productivity gains. So like any business that is for profit, you’re trying to increase your profitability and you want to maintain or increase your competitive advantage. To do that, you have to year over year increase your net productivity, your productivity gains have to grow faster than the rate at which your costs are growing. Now in a low inflationary period of time, that’s easier to do in a higher inflationary time, it’s incrementally more difficult to do increase productivity faster than the pace of costs.

(06:32):

Why it’s always been there for any for-profit company, I just think it’s now increasingly important and increasing focal point. You hear many professionals, you hear Zuckerberg with meta talking about how this is the year of efficiency and for us, we’re leaning into that as well. It’s the year of relentless efficiency and I know many businesses are focusing on that. So that was the big why for us is this focus on productivity. And then again, how many tools in the toolbox do you have to increase productivity at? I mean, inflation is what, 6%? How do you increase productivity at 6, 7, 8, 10% a year? That’s a difficult challenge for any business and automation is a pretty serious tool in the toolbox to help you achieve that.

Hani Batla (07:23):

Yeah, so same thing, right? The dove starts with the why we’re doing it. I am going to repeat it. So yes, it is a year of efficiency. Us like every other retailer out there is looking for efficiencies because we are in a environment where we are constraining our costs and that’s really one of the big drivers. But when we think about automation, it’s just to look at from the lens of cost saving is not the right path. We have looked at it from cost saving, sure, but that’s a side effect of what we’re doing. The real why for us is about getting more value and leading to outcomes that are positive for our customers, that are positive for our business that lead to revenue growth. And if as long as your core belief is one of positive and adding value and growing or making things better, then I think that’s the right path and that’s been our north star for using automation and AI for that purpose.

Jared Shaner (08:25):

Yeah, it makes a lot of sense. I mean we’re constantly asked ourselves, how can I be able to optimize my business? And the idea that there’s always going to be a one fit size solution out there as you combine has proved itself to be a little bit inaccurate even if you’re using some pretty powerful systems. And I’d love to dig in a little bit more specifically on some of the things you guys have implemented and sort of that journey. So what are a couple things, I guess it sounds a little bit more on the automation side and the AI side that you’ve implemented as examples. Sure.

Measuring AI impact and tangible results

Hani Batla (09:05):

I’ll take one example. I mean we have several use cases we’re looking at opportunities for using it, but my favorite one recently is aam. Again, we’re into cameras. One of our fastest growing parts of our business is our trade used business. And that one’s an interesting one because in an environment where creators need to upskill their gear or they’re rising up the ranks from being amateurs to professionals, they can’t always afford the most expensive gear out there because the cameras and the lenses they all help, the better they are, the better your art is. But we found that we had a broken experience there for our customers where customers would come to us, they would want to trade with us or sell their gear to us, and we found that given who they interacted with in terms of a human being, a solicitor who’s working with them to price out their gear for them so that they have a fair price so they could sell it to us and then after that they could either buy new gear or whatever.

(10:02):

But bottom line, we found that the huge pain point was we were starting to see very negative reviews across the board about that experience. And so we started to think about what is it that we can do to make that a fair more uniform experience for everybody who comes to aama? And so we started this thinking of what if we optimize the human solicitor? Yes, you’ll still need humans in the back end, but really what you want to be able to present to a customer is a fair price. One that’s based on all sorts of data points well beyond the capability of a human being, right? You look at things like inventory, you look at demand, you look at competitive information, right? And what we tried to optimize for, we laid out a machine learning AI project where it was all about coming up with the right algorithms to predict first the right sell price because that tells you how much a product is worth something in the market and sell it for the right price and be competitive in that price.

(11:01):

There are other companies who do this, but we want to beat them as well and take more market share. But then more importantly as you have that, you work backwards from that to getting to the perfect buy price. And once you get to that sweet spot, you actually are more honest with what you can offer. You can pull the levers to take your margin down more or anything the machine learns constantly. You’re no longer dependent on a human being or two different human beings giving two different experiences. Now when you’ve uniform brought value there and you’ve upped the customer experience, that is where we see customers then respond more positively and we, it’s still very much in almost rolled out phase, but we have been testing it. It is efficient and it is causing efficiencies inside our company because now those very people who were dealing with quotes all day long don’t have to do that. They are freed up to do more higher value stuff. They can go back and refine estimates and make even better offers if they need to. But the real focus, again, like I said earlier on the North Star was about the experience and we wanted to optimize and that’s helped us. So that’s just one use case of what we are doing right now. Absolutely. Amazing.

Antonio Colicchio (12:15):

Love that. I love the focus that Haney’s been talking about in terms of the attention to the customer and the customer experience. So I didn’t want to discount that, and that’s a big why. So thank you for mentioning that and that’s an awesome example for us. The automation is really about specifically robotic process automation, RPA does Anybody here, just quick raise of hands, are you familiar with RPA or have any experience with RPA? Raise your hands if you are not many hands. So let me give a quick example of what that means then to kind of make it a little bit more tangible for you. So when I’m talking to my buddies back home and they’re like, what the heck is RPA? The easiest way I could explain to them is that most people know what a macro is. You may have built a macro, I’ve been familiar with a macro saying in XL or something like that, and I’m like, it’s a macro on steroids.

(13:09):

And they’re like, well, what do you mean? Tell me more. So we’re a retailer. I’m sure everybody here, if you’re not in retail, you can relate to retail. I’ll give you a specific example of a bot that we’ve built, but then I want to talk about two other bots that touch on the concept of the customer experience. I love that note and the associate experience. So we pay Abercrombie Fitch, we will pay carriers to deliver our stuff. So you place an order and then X, Y, Z carrier is going to attempt to deliver it to you. If they don’t make that delivery, we can file a claim against the carrier to get our money back either for the shipping or the value of the product up to a certain amount. Not to get into the details, but you can file a claim and you can get money back.

(13:51):

What does it take to file a claim? Well, you can hire somebody and they can go to the appropriate website, download the form, and now they got to populate the form. How do they do that? They got to go to system source, data, X, y, Z, grab various elements of metadata and information, populate the form, and then they have to process, submit the form. Then maybe they got to track it to see if there’s payment that comes back or it doesn’t come back. That is an example of a process that we completely automated. We automated that across multiple carriers, so hopefully that makes it a little bit more real in terms of what we’ve done. I want to talk about two examples of things that we focused on that we’ve built. We’ve been doing this for, and I should say, so we have a whole center of excellence around this.

(14:38):

So we have a method and a process to intake ideas, vet ideas, and then take ’em to development and produce bots. We’ve been doing it now for a couple of years. Formally we started very scrappy and informal, but it’s been two years of a formal journey. We’ve produced a couple of dozen bots across the entire business. I just gave you an example with carrier claims. Two other examples, which I think are exciting, which touch on the value and the benefits you can get because when you hear an example like that, you might immediately think, wow, this can reduce costs and it can help me increase revenue. Yes, but there’s other benefits. We’ve built bots that have directly impacted the customer experience. I’ll give you an example. So one of my responsibilities is customer care and pretty much every retailer holiday season is the most important time of the year.

(15:31):

So we will do promotions at this time of the year where we will guarantee delivery for Christmas. If you place an order by a certain date, you’ve all likely experienced this. Well, inevitably, yes, it’s a guarantee, but guess what? It’s not a perfect world. Stuff breaks in different areas of the supply chain or you got weather stuff happens and a small percentage of those will miss delivery. We won’t make it. We just promised you that we would. That’s a terrible customer experience, not a way to build brand loyalty. So in the past, what you would do is, well, if we missed the Christmas delivery, you as a customer would call my team the day after Christmas and let me know that you missed the delivery and you ruined my Christmas. Got to make it right and we would make it right. Of course, we’ve been increasingly moving our customer care strategy from a very reactionary approach like the one I just gave you to proactive.

(16:27):

So in the last few years, rather than wait for it to happen and then you tell me that it happened, we’re telling you it’s going to happen ahead of time. So we have people process tools to be able to say, well, a couple different scenarios, we are likely going to miss it. It’s not looking good. And then we have a path to inform you about that ahead of time and take you through that journey of it’s not looking good. If it doesn’t look good, what do we do after? Or you know what? It’s not going to happen. We’re totally going to miss, but we’re telling you about it ahead of time. And that was, we did it for a couple of years, which is scrappy people process, and I had people working throughout that entire holiday weekend, which sucks from an associate perspective that you’re working Christmas Eve, Christmas day, the day after to do this thing, which you’re doing it for the right reason, which is take care of the customer and do the right thing by the customer, but not the best associate experience. That’s an example of another process that we’ve mostly automated through RPA. Now the customer benefits because they’re getting more efficient, more effective information quicker, sooner, and then the associate benefits because they can enjoy time, spend some time with their family, had a couple other examples, but I’m sake of time. That’s great

Hani Batla (17:42):

Example. We did the same thing. I know what you’re talking about. We did this crappy, got a guy out here in the audience who does exactly that right now, him and his team came up with the solution to do exactly that, but that’s exactly where we’re going to. We are moving towards more automated proactive communication if you tell them in advance, and customers are more tolerant now for those kinds of things, but automation helps there.

Stories of successful AI implementations

Jared Shaner (18:02):

It’s clear that you guys are doing a lot of the tinkering, so I’m glad that that was in a lie on my behalf. I think it’s funny, I work in technology, but maybe I’m a little bit jaded. I try not to follow the trends as much. So AI is a hot button thing, and it sounds like you’re doing some really cool things. I’m interested in the process as you guys implemented. Were you setting KPIs or are there ways in which you’re measuring success to say, Hey, let’s keep on tinkering further with this thing versus, okay, this was a good idea, but it’s going to be too hard to really get to where we need to go, or maybe it’s not going to work at all. It didn’t work how I thought.

Hani Batla (18:44):

Yeah, and no, because there’s things you can measure today. So for example, the used trade business example, that’s a measurable thing. You can measure the efficiencies gain through implementing the AI and ai. You can measure it as you tweak it, you improve it, you constantly mess around with it to get a better, more positive outcome. It’ll reflect in everything from customer reviews to whatever. So that totally trackable. It’s the other areas where we’re starting to now think about using AI or RPA for that matter where it’s about, I’ll give you an example. In our finance team right now, we had an individual who literally wanted to solve for a function that he was doing and it was consuming an immense amount of time. He was taking invoices from thousands of vendors, putting it all in, trying to get it all processed and all that stuff.

(19:35):

And actually, I’m going to quote Antonio here. He used an amazing term citizen automation. So it was actually an example of that, and I’m going to give credit to Antonio here for that term. So this guy was exercising his right as a citizen of the organization to do citizen automation to improve a process free up some time. Now, how do you quantify that right now? You could argue that he can now because he no longer has to do this function and he saves two hours, right? It’s very hard to continue to quantify. Some of these things start with a simple belief that someone felt the need to make a change to how they do their job. They believe they’re going to be more efficient, and so you got to go with that. So you can’t always back it up with tracking it, having the metrics defined. I think once we have it going and then we take on that next citizen automation project or whatever, I think we’ll get better at learning how to measure it and have some KPIs around it. That’s how I’m thinking about it.

Antonio Colicchio (20:30):

I love it. I’ll kind of piggyback on that. So I love to start with the belief that’s how we started the journey as well. Well, we talked about the why, right? That was the motivator. But you can’t go to your CFO and be like, Hey, give me this amount of money. I want to do this whole center of thing or on RPA because I think it could be worth X, y, Z. Again, the individual that talked about strategy had a lot of good insights into how to approach something like that. But for us, what we did is we started scrappy. We knew there was value and we knew it could be a really powerful tool in the toolbox against the Y that I talked about earlier. So what we did is we basically did a proof of concept. We started with a bot and appropriately it was to fight chargebacks.

(21:16):

So anybody here is fighting chargebacks. So we built a bot that was the first bot we built to go out and automate and fight every chargeback. So we did that in a very scrappy way and we had great business partners that came to the table with us and said, Hey, we want some skin in the game here and we want to prove and show that you can do it and we can do it together. Then maybe we can continue to end this journey together. So we essentially did that for pretty much free and we took it to the C ffo and said, Hey, look, we did this thing in a month. We built this thing in a month and look what it’s bringing in. And he was like, cool, go do more. And that led to a more formal center of excellence and a whole process around how do we intake ideas across the business, how do we vet them, how do we prioritize them, how do we take ’em to development, et cetera.

(22:05):

On your question around though, as it relates to business case and return on investment, how do you think about value? What I’ve learned in this journey, I’ve been doing it for a couple, two, three years now, is when I talked to others that do automation specifically RPA. We just went to a conference late last year and when you talk to people in the breakout sessions almost exclusively, everybody talks about the benefit in terms of hours. Touched on it a little bit, but he also has a broader sense of the benefits. But in this space, a lot of companies and individuals talk about we saved 10,000 hours or 20,000 hours or a hundred thousand hours, and that’s how they’re calculating the benefit. And I think that’s great and you got to do that, but we’re pretty fortunate when we went to the CFO and we went back to him and said, Hey, we want to formalize this, we acknowledge that and we all face this, there’s a finite amount of resources, the money, the bank account’s not just open.

(23:03):

So you have to prioritize your investments, of course, you want to prioritize them in terms of the biggest return on investment, the things that you can quantify directly like point to a place in the p and l and show me where the revenue went up or the expense went down. That’s a direct benefit that if you’re vetting and prioritizing 50 ideas across the business, what’s inevitably going to happen In most organizations, all of those ideas float to the top and all the other ones, which certainly have value in a different form, they float to the bottom. So we did this really great thing, and again with a very open MI to cfo, they were like, look, we don’t want to do that. We know there’s different benefits in different buckets and kind of had a handshake deal that we would have what we call the concept was effectively a well diversified portfolio talking to the C FFO and CFO speak a well diversified portfolio of benefits.

(24:01):

And for us it’s in the following buckets. We said that we would diversify the buckets. The first bucket is direct benefit, and that is like I just alluded to you, point to a place in the p and l. Look, dollars went up or revenue went up or expenses went down, and we said, that’s going to probably be the majority of what’s going to drive value when we vet these ideas. And he’s like, yeah, that makes sense, but let’s not forget the other buckets. Then there’s indirect benefit. Now an indirect benefit is essentially bots that create capacity, and that is we may have automated a process or a part of a process that say saved a thousand hours. Well, that’s not an FTE, so you don’t reduce your headcount anywhere for a thousand hours, but that thousand hours has value. Now the associates on that team were able to take that time and distort it to something that they weren’t doing before or do something that they were do, but do it better, faster, cheaper. So that unlocked value. And then the third bucket, which I don’t want to minimize is risk mitigation and just quality improvements. We’ve built bots for our legal team that help us to identify and categorize intellectual property so they can have faster access to it in the event of litigation. Now, how do you put a dollar on that if you were vetting stuff in this pipeline of direct benefits? So hopefully that gives you an example how we thought about it that’s much broader than just hours or direct quantifiable benefits.

Learning from AI failures

Jared Shaner (25:36):

Yeah, well, I love hearing there is sort of some of the advice around taking iterative improvements or taking small steps to try to prove it. Give me a dollar, I’ll come back with $2 babies, and that sort of grows and eventually you end up with a center of excellence like you have Antonio. I’ll share a quick story on automation. So hopefully it’ll open up us all to be a little bit more open to share our failures. If you visit our website right now, there’ll be an ugly mug of me that pops up. I thought that there was this great idea to use an automation because one of our problems is our capacity to answer people when they want to talk to us. And I’m a service provider, so when someone wants me, I have to fly in and make sure I’m there because otherwise Antonio might never reach out again on our little website.

(26:28):

What I discovered is I gave the ability for people to schedule. Unfortunately, even today what I discovered was about 14 different people blocking me out throughout the entire day to try to sell me things and then also try to convince me to offshore all of my development. So that was a bit of a swing in the miss. I’m going to keep on working on that as my iteration. I’m curious if there’s any small failures. I’m sure you didn’t swing in the miss as bad as I did that might be able to help those in the crowd today as they think about their journey with AI automation.

Antonio Colicchio (27:05):

Yeah, I can jump on that one, honey, while you’re thinking about that. Yeah, we had a swing and a miss, and again, this goes back to I just kind of the awareness we had going into this with our executive leadership team is like, this is still kind of new to us. We had some early wins, but we were like, we’re still not entirely sure how to do this, right? So we acknowledged that we’d have some missteps and we did. So there was an example where we, and we learned, I don’t want to take the wisdom of the previous, go back to the strategy conversation. It really resonated with me. What he was saying is you learn most from your failures. And that is certainly true and I know we all have experience with that and believe it. But to keep it short, we essentially, again, I was alluding to this process we have for intaking ideas.

(28:05):

So we have a process where we outreach to the business, they know who we are, how we do it, and why we’re doing it. So we have that awareness across the business, you got an idea you want to automate, come to us and we’ll talk about it. And so we have a whole process for filtering through that and then getting to a business case and then taking it to development and everything. It was a change management process, a commitment process where before we kick off the bot, we have the executive sponsor there, the subject matter experts, a business partner from finance and human resources. We all hold hands. We say this is the process and this is the part we’re going to automate. This is what the value is and it’s going to be great. So are we ready to do this? Are we committed?

(28:46):

Are we committing? Yes, we’re committing. It’s kind of like getting married, the commitment process, like, great, let’s do it. Needless to say, then we spent weeks building a bot that on the day it was going to go live, we realized that we didn’t talk to all of the right subject matter experts and that the process that was being implemented was essentially obsolete immediately because there was another change happening in the business that rendered it obsolete. It was going to render it obsolete within weeks. And that’s a pretty painful failure, right? Because you thought you had all the right people engaged, you thought you were doing it in the right way. We thought we were talking to the right people, we built it. So we spent the money and then it was obsolete on day one. But we learned a lot in terms of how to re-engineer that process, the funnel, the intake, the commitment process, the outreach, make sure we’re talking to the right people at the right place at the right time to mitigate obviously the risk of that occurring in the future.

Hani Batla (29:49):

So we haven’t been doing this long enough to have any Mrs. Thank God. But I’ll tell

Jared Shaner (29:53):

You, it’s the genius leadership I think. Yeah, I

Hani Batla (29:55):

About that.

(29:57):

Good luck maybe. No, but I will say I think where I see the greatest risk for us to have misses is misses of expectation. I think because this is still so new, I think it takes a lot of time and energy to educate other leaders within our organization about why we’re doing this and whether it’s even worth doing it. I’ll give you a prayer. For example, a few months ago I brought up this idea of like, oh, we should be doing this and this and this, and in the immediate response I got was like, I dunno if you have the capacity to do this, but no, I’m like, this is good. We can do this at very nominal costs. We have the tools. But that’s the challenge. I think that’s one, people may not understand it because it’s still fairly new and they don’t want to bless the initiative in the first place. Thankfully we did it anyway. And then second, when you start doing it, people start having very unrealistic expectation of what this could result in the savings, the efficiency gain. Oh, what do you mean you can’t have one developer do the job of 10 now with this kind of intelligence AI in place, right? That’s where the risk is. And I think that’s where we’re still kind of being cautious as we’re doing this. So we don’t miss that. Let that hurt us.

Jared Shaner (31:09):

Thanks for sharing. I know it’s hard to share our failures. I do have a couple more questions, but I want to make sure I don’t rob the audience if they have any questions for anyone here.

Future trends: generative AI and automation 

stuart Gold (31:21):

Alright, if you have a question, raise your hand. I have the microphone. Anybody. Okay, let’s keep going.

Jared Shaner (31:30):

Awesome. Let’s talk about the future then. We talked about where we came from. What started your journey, even share a little bit of the failures other than chat, GPT and the machines taking our jobs individually as well as for our industry as a whole. What do you think’s on the roadmap and what’s the future of the next three to five years?

Hani Batla (31:57):

So I know there was a wonderful session about chat GPT, and we learned a lot of history about it, but some things I even didn’t know. But here’s how I look at it, at least from our take. The whole role of generative AI in various functions across our org is a very real thing. The first thing I did when Microsoft, at least it’s bing chat with chat GPT-4 on the backend and all Sydney and all that stuff. I asked, the first thing I asked was, what’s the best camera store for what’s the best camera store for professional cameras in the us? And immediately the answer was a good one. We were in there, we were number two, wasn’t happy about that. Again, the line said another Canberra store in New York City, which I wasn’t happy about either. But overall it was still, we made the cut off the five that the AI thought were trusted enough to buy professional can store.

(32:51):

The part that scared me was the very next prompt that was underneath that, which was, which one do you recommend? And that’s what my team and I have been thinking about. Now, how do you prepare for the day when you ask that question, when anybody on the internet asks that question, which one do you recommend? And what will the answer be? And it’s going to be about trust. It’s about in the past we marketed to human beings, we optimize our SEO so people could see those links come up and our name shows up. But how do you optimize for AI and how do you get the AI to trust you? So that’s where we’re really thinking about, thankfully, just so that everybody knows what the answer was the first time was it actually did try to answer it, but it said, well, what do you want?

(33:39):

Mirrorless DSLI totally missed the question. Now if you click that same prompt, I do the same search every single day on Bing and on board, and now it’ll say, well, there’s a lot of factors that get into that you need to take in. There’s pros and cons for each one, but if you want this, then you go to Arama. If you want that, you go to this, it won’t give you a definitive advancement, but one day it may just do that. So that’s what we’re thinking about the future of that plus the value of using generative AI in unlocking more productivity. So not 10 XA developer, because to me, a developer is an artist and he’s creating something new and solving a problem. But office workers, I think the previous guys talked about it, right? There’s going to be guys whose jobs won’t be necessary because Microsoft Office with copilot could now probably replace at least three people of a team of five because those two guys can be way more efficient with their time and they can utilize their creativity and productivity and take it up 10 x. So that’s how we think about it. And we’re thinking of efficiencies in the workplace and then the future of how as retailers, the world will look at us.

Antonio Colicchio (34:49):

If I have time, do I have time? Do you want me to get into this? I’ll do the short answer. My thought here is I don’t disagree with what anything Ani said, and I think the AI presentation touched on this. I thought that was really great. I think the simplest way to think about is I see a convergence of innovations that are going to be coming together that are and will be continuing to mature and come together over the next several years. The AI session talked about it, speech recognition, continued developments in speech recognition, continued developments in image recognition, the proliferation and maturation of open APIs, and then from A RPA and a process or automation perspective. Process mining and process mining is the concept of you put something on a desktop and it basically monitors, clicks and screens of the individual on that machine.

(35:47):

And it’s smart enough to know capture trends, spot trends, and see repetitive tasks and then suggest to you automations that’s not far from the realm of reality already. So when you put a smart element like AI in the mix of those things, you put an AI brain and you tell it. These are the tools that your disposal, speech recognition, image recognition, you’ve got all these APIs at your disposal and you’ve got this process mining thing that tells you, look, here’s all the repetitive tasks that are happening across this organization and who does it. And if person A works with person B, with person C to D, they’re all in a, they might have a process within themselves, but across themselves. I think that’s when you start to see some pretty sophisticated things like the AI being able to spot a trend in any process you can imagine and either suggest or really start an automation process.

Closing remarks

Jared Shaner (36:52):

Amazing. I look forward to continuing this conversation after the session. If any of you want to catch up with my better haves or two thirds here, and Antonio, we are on the rest of the day. I really appreciate you making time for us.