KIT Episode Episode 210 - Robots, Industry Finance, and a Practical Approach to AI

[00:00:01] IT Businesses in Sacramento, California.

[00:00:03] From somewhere deep in the cloud and the corners of the earth, this is The Killing It Podcast, with a focus on helping you make sense and dollars of all things IT. With your hosts, Dave Sobel, Ryan Morris, and Karl Palachuk.

[00:00:21] We're back! Welcome everybody to Episode 210 of The Killing It Podcast. Oh, you guys are in good voice today.

[00:00:34] Well, you know, it comes back naturally, particularly when we warm up for a little while, secret listeners.

[00:00:41] We're all back together again though, with cool stuff to discuss because things always fall into our laps in terms of opportunities, particularly when we want to follow up on stuff.

[00:00:49] And the 15 minutes of warm up didn't hurt.

[00:00:53] It did not! We're all well and limber. But I'm going to start off with something fun, gents.

[00:00:57] As we think about the end of the year, when no one's around to judge you, do you ever play Christmas music before Thanksgiving?

[00:01:04] I will say 100%. Once my daughter moved out, which is now like nine years ago, I play Christmas music whenever I want to.

[00:01:14] And in particular, any time when I'm in the fall, I go to Australia and in October I can buy Christmas Aloha shirts. And I do every single year.

[00:01:27] See, I will say absolutely positively never, but I've surrendered being a curmudgeon and being angry at the rest of the world for doing it.

[00:01:39] Right? Like we live in a reality that this is going to happen. Everybody's going to go directly from Halloween to Christmas and all of all of that's going to happen.

[00:01:48] And for decades, I've been that guy over like tilting against windmills saying, no, it's too early. You can't do that.

[00:01:57] And I've admitted, you know what? If people enjoy it, enjoy it. Enjoy what you like, when you like.

[00:02:02] But no, I will. You will never catch me watching a Christmas movie or listening to a Christmas song prior to Thanksgiving.

[00:02:13] I really don't want to yuck other people's yum. Right. So for me, my wife and I, like we get our Christmas tree either the weekend after Thanksgiving or the first sort of weekend.

[00:02:25] And for me, that's the beginning, right? We get the tree, then I start playing the music. Like that is just the transition. That's what works for us.

[00:02:33] I've also softened on a list and I mean this sort of like in just the, I want to understand. So my mom's got a little bit of, she's got some dementia, some short term memory stuff.

[00:02:42] She's taken to keeping her Christmas decorations up year round because they give her joy.

[00:02:48] And you know what? Again, I'm not going to yuck someone else's yum. If she gets joy out of that in April, you know what? That's okay.

[00:02:57] Precisely.

[00:02:58] That is totally okay. So it's, I think for, it's about embracing the tradition and I'm really trying, like I used to be like sort of like, ah, not too early, not too early because I'm not ready.

[00:03:09] And now it's more than like, you do you people. Exactly. Exactly.

[00:03:14] I would have my decorations up now, except I'm having a get together on the weekend.

[00:03:19] And I know that I will just get so much garbage, like from people who are like, where are your decorations up?

[00:03:26] See, the social pressure. Someone's yucking you yum.

[00:03:29] I do have the music on.

[00:03:30] So quick note, five years ago when we started this podcast, we were literally watching all three of us on our other monitor, Notre Dame burn.

[00:03:41] And Notre Dame is about to reopen and Macron is making a big deal.

[00:03:46] I promised five years and it'll be less than five years.

[00:03:49] And it's, it's amazing.

[00:03:51] It's not done yet, obviously, but it just, just a note sort of historically.

[00:03:57] Time flies.

[00:03:58] Yes, time flies.

[00:03:59] And cosmically, it's a little sad that they weren't able to get it done in time for the Olympics in Paris, which as we all watched through the summer, that was brilliant.

[00:04:11] They did a fantastic job.

[00:04:12] The leveraging of all the monuments and the famous locations as sporting venues.

[00:04:18] If they could have said, oh, and by the way, allow me to play you some bells.

[00:04:23] That would have been epic.

[00:04:25] But the fact that they're doing it in less than five years is actually logistically unbelievable.

[00:04:32] So kudos to them and a reason to take a trip.

[00:04:37] Exactly.

[00:04:38] Alrighty.

[00:04:39] So our first topic today is, unfortunately, robots.

[00:04:44] None of us actually want to talk about robots.

[00:04:47] But I'm just, I keep asking myself, where are they?

[00:04:53] And to be honest, the pandemic has sort of slowed us down a bit.

[00:04:56] And then there was a bit of a backlash as people said, well, you know what?

[00:05:00] But we'll just lower the cost of labor to the point where human beings will do this.

[00:05:06] We can't afford $15,000, $20,000 robots.

[00:05:09] And so there's lots of reasons for the delay.

[00:05:12] But time has marched on.

[00:05:15] We've had a standing bet for five years.

[00:05:18] Who will have our burrito or pizza delivered by robot?

[00:05:22] Five years later, no one has won that bet.

[00:05:26] It's just like, okay, is this all fake?

[00:05:29] What's going on?

[00:05:30] On the back end, robots are in manufacturing and they're expanding there.

[00:05:35] But I've been to restaurants with robots and humans that work spectacularly well.

[00:05:41] And then I've been to restaurants where I still have to wait 45 minutes for a table

[00:05:46] because they don't have enough employees.

[00:05:49] Where are my robots?

[00:05:52] And will it ever end?

[00:05:55] I think it's an economic problem, right?

[00:05:57] Because we're either trying to build these all-purpose robots that do everything.

[00:06:02] And it's like, well, the math on that is just impossible, right?

[00:06:04] Are you going to be able to...

[00:06:05] It's a complicated technical challenge.

[00:06:07] We already know that the algorithms, and I'm using that broadly because I'm not trying

[00:06:12] to use AI in the first 10 seconds, but the whole bit of that is expensive.

[00:06:16] And by the way, it's getting more expensive, not less expensive.

[00:06:19] So the economics of a multi-purpose robot are just brutal.

[00:06:22] And I actually think we're also sort of dismissive of the small utility ones that we have.

[00:06:28] I mean, I was thinking about this in prepping.

[00:06:31] Like, I have a Roomba, right?

[00:06:32] Like, every day.

[00:06:33] Like, it does its job, and it's totally part of the background.

[00:06:37] It is a special-purpose, solves-one-problem robot, and it's not particularly glamorous, right?

[00:06:43] Because it's sweeping, right?

[00:06:45] Like, it's...

[00:06:45] But it's real utility.

[00:06:47] I mean, and I'll say, like, for all the techies, my wife chose to buy the Roomba.

[00:06:52] Like, it was one of the few things that we have in our house that I did not push for from a technology perspective.

[00:06:58] She said, I want a Roomba to clean up the cat litter.

[00:07:01] Like, it was literally the...

[00:07:02] I identified the problem, and I have a solution to that.

[00:07:04] And it is exactly the buying journey you want, but came from the person you wouldn't have expected in my relationship.

[00:07:10] And so...

[00:07:10] But it's not glamorous, right?

[00:07:12] It's not a glamorous, like, cool, we'll serve me drinks, be my butler, do all the things in my house.

[00:07:17] Nope, sweeps my floors.

[00:07:19] Right?

[00:07:19] And so where I think what we're struggling with is utility, right?

[00:07:24] The balance of utility versus cost.

[00:07:26] The economics here is just sort of still broken.

[00:07:28] I wouldn't be surprised if there hits a moment where that floodgate opens up and something happens.

[00:07:35] And like most of these kinds of things, we've put it wrong.

[00:07:38] Well, and I think...

[00:07:39] So I'll say I buy that there is an economic problem, as you've described.

[00:07:45] But I think what we're having is more of a storytelling problem.

[00:07:49] It's less about whether the technology can be useful.

[00:07:53] It's more about what our expectations are based on all the stories that we have been told.

[00:08:00] What we want, what we say would be utility for us is a fully functional humanoid looking object that is both intelligent and capable and has all of the abilities to do all of the things that I would want a personal assistant to do.

[00:08:18] And then we realize, yeah, the only way that that's actually working right now is apparently at the launch events.

[00:08:26] We put people in the crowd and we control them with video game controllers, right?

[00:08:32] That's the only way that we're getting anywhere near that anytime soon.

[00:08:37] But we go back to the functionality of what is available and what is necessary.

[00:08:44] And the Roomba does a great job.

[00:08:46] And that restaurant delivery robot that brings you your order while the staff person is off serving a different table, that works really well.

[00:08:55] But they're not glamorous.

[00:08:57] They're not cool.

[00:08:58] They're not the extra level of storytelling that we've come to expect.

[00:09:03] And we've gotten ourselves to this level of I see the technology.

[00:09:08] I recognize that it works.

[00:09:10] I've built up radically unrealistic expectations.

[00:09:14] And now I'm not happy going back to square two and making the incremental progress through all of these things.

[00:09:22] I will say travel around Asia and you will find, having been over there a handful of times in the past year, restaurants, hotels, offices, subway stations.

[00:09:37] Robots are there.

[00:09:38] They are moving around.

[00:09:40] Most of them look like little mini fridges with wheels on them.

[00:09:43] But they're doing stuff.

[00:09:46] And whether they are cleaning or delivering or answering questions or they are redirecting people to shorten a line at the subway station in Japan, that was freaking brilliant.

[00:09:58] I was like, oh, look, we're all being shepherded by this border collie that's actually a robot.

[00:10:05] And it worked out just fine.

[00:10:07] We all went, oh, the machine told me what to do.

[00:10:10] And we did it, right?

[00:10:11] Like, there are use cases out there.

[00:10:14] I think our problem is we expect them to be so advanced and the reality is much less advanced.

[00:10:22] And so I think most of us are just like, well, that's not fancy enough.

[00:10:25] I don't want to participate in version one, two, and three.

[00:10:29] I want version infinity.

[00:10:31] Well, I don't need it to look like a human.

[00:10:33] I will say that.

[00:10:34] But a couple of things.

[00:10:35] First of all, why doesn't Roomba make something to mow my lawn?

[00:10:40] That will save me.

[00:10:42] Don't they?

[00:10:42] I thought they did.

[00:10:43] I thought they did.

[00:10:44] I thought they did.

[00:10:45] Why didn't it work?

[00:10:47] Like, why doesn't everybody own one, right?

[00:10:50] Or to sweep the sidewalks or whatever.

[00:10:53] The second thing is, in the pandemic, I thought that was the opportunity for Amazon to figure out delivering packages by drone or by robot.

[00:11:05] And we'd all have little drone stations so it would be securely dropped off.

[00:11:09] Instead, they bought thousands and thousands of really big trucks.

[00:11:16] And then they outsourced delivery.

[00:11:18] Like, they created an entire sub-industry to compete with UPS, for goodness sake, rather than investing in robots.

[00:11:27] So Amazon uses them in their warehouse.

[00:11:30] They don't use them in public.

[00:11:31] Is there a public piece of this?

[00:11:34] And we have one minute left.

[00:11:37] By the way, just so I've been accurate.

[00:11:40] iRobot does not, so the company makes Roomba does not make one.

[00:11:43] There are other manufacturers that make it.

[00:11:46] All right.

[00:11:46] Very good.

[00:11:47] They can mow your lawn for you.

[00:11:49] See, I think you're right, Carl.

[00:11:51] I think the first acceleration, the real acceleration is going to come on the industrial and manufacturing side, inside warehouses, inside the shop floor.

[00:12:01] I think there's going to be a lot of use cases there.

[00:12:04] But by the time it gets to us, see, I was ready to gloat.

[00:12:07] A couple of years ago, they announced Walmart had one of their delivery beta test sites literally a couple of miles from my new place in Utah.

[00:12:16] And it was right there.

[00:12:18] And I was like, ooh, I'm getting the burrito delivered.

[00:12:21] It's going to happen.

[00:12:22] A couple of months ago, they announced that they were shutting it down because it wasn't because it didn't work.

[00:12:28] It was because they weren't getting the right consumption models.

[00:12:33] They weren't able to deliver large enough orders to make enough money to make it worthwhile.

[00:12:39] So, you know, I was ready to gloat.

[00:12:41] And then I was stymied.

[00:12:43] So we will see where that comes from.

[00:12:45] All right, guys, let's talk about our second topic.

[00:12:48] And this one is focused on our own industry and some of the players that are coming into our space.

[00:12:55] In just the last handful of years, we've seen the arrival of or the growth of some billion-dollar companies in the technology channel space, right?

[00:13:06] Whether focused on PRM applications or MSP standard applications, right?

[00:13:13] We've seen an awful lot of these companies that are growing up and becoming very large.

[00:13:19] Trouble is they're not here to service necessarily everybody based on what we can see.

[00:13:25] They're focused on, as most PE-based companies are, the very large players, the top handful of percent of solution providers and MSPs in the industry.

[00:13:35] And they're going to provide a tremendous layer of service.

[00:13:38] This feels like a new chapter in the economics of being in the business of a technology service provider company.

[00:13:49] What do you guys think about this evolution?

[00:13:52] Are you seeing any good news come from these large players?

[00:13:56] Are we afraid of them usurping and consolidating the entire space?

[00:14:02] What do you guys think?

[00:14:03] Oh, Ryan, you almost said business of tech.

[00:14:05] You almost did it.

[00:14:10] So, you know, I'm balanced on this one because I'm still debating in my own head with the way I'm thinking.

[00:14:16] But for listeners, here's what's going on in Dave's head.

[00:14:18] So I keep looking and saying IT spend continues to grow, right?

[00:14:22] If I look at the history of our industry, it's been pretty consistent that IT spend grows.

[00:14:28] So in theory, a continuously growing market has more room for more players.

[00:14:33] And additionally, as it gets, the numbers get larger and larger, there's also a space for larger players to scrape off larger bits.

[00:14:43] Because, you know, that 10% growth in early days in raw dollars, 10% now is significantly more money.

[00:14:52] And as all companies have kind of, we've got levels of growth along that there's more room to play.

[00:14:58] And I continue to think like, like, like.

[00:15:00] So large companies like large companies, mid-sized companies like mid-sized companies, small like small.

[00:15:05] And as more large companies become more technology enabled, it feels natural that there's just more space for these players.

[00:15:13] So I, at first glance, don't say, don't think this is necessarily bad.

[00:15:19] What I do also look at is, is I think we're going through another one of those swings where some of the players within the space that have served like customers,

[00:15:29] that have typically small, you know, if I do my small like small, well, some of the vendors that serve small, particularly in the technology space,

[00:15:36] have become mid-sized and moving into large companies, while their customers remain small.

[00:15:42] Right?

[00:15:43] And what I think we're seeing, particularly as this move from integration ecosystems to platform plays, as I think about the MSP space,

[00:15:51] it's like, oh, it's the pendulum is swinging because the larger companies have figured out that the best way to serve the smaller ones is to convert into a platform

[00:16:00] and do that large company serving small company at volume play.

[00:16:05] Like, that makes more sense.

[00:16:06] The less interconnected tendrils of integration I got to deal with, the better it is for me to serve.

[00:16:11] So I think the pendulum is swinging because of that.

[00:16:14] But I'm not necessarily worried about like, oh, this is going to break, you know, like we're just going to run out of money or we're going to run out of economy or service areas.

[00:16:23] I do think there's certainly nothing wrong with big companies.

[00:16:27] And I'm a big fan of money.

[00:16:28] You know, if you have any spare money sitting around, spending my way.

[00:16:33] There is a change in the ecosystem.

[00:16:36] You're right.

[00:16:36] That when you have a group of people at the top, a group of companies at the top that must get, and 10% is quite low, I would say.

[00:16:45] They want 20%.

[00:16:46] They want 40%.

[00:16:48] And when you just arbitrarily say, last year, everybody, every company we owned had to bring in 20% profit.

[00:16:56] Next year, it's 21.

[00:16:57] The year after that, it's 22.

[00:16:59] Right.

[00:17:00] For no reason, no rules, no, it's just arbitrary.

[00:17:03] That's going to have an effect for those big companies.

[00:17:07] But at the same time, literally 99% of the companies in our space have 10 or fewer employees and 95 have four and fewer employees.

[00:17:19] And that is where the growth comes from.

[00:17:22] When people are 18, they don't get out of college or high school and say, oh, I'm going to start a billion dollar company or a hundred million dollar company or a one million dollar company.

[00:17:34] They say, I'm going to go into tech support and I'm going to figure out what I want to do.

[00:17:40] And I think we're in an era where all of the money for that group of people has been sucked up.

[00:17:48] And it needs to be replenished somehow.

[00:17:51] We need to be able to support them and bring them up and have some standards.

[00:17:57] Right.

[00:17:58] Now, right now, Dave, what you alluded to is it's almost kind of like a Salesforce model.

[00:18:02] You can go and say, I need just enough RMM to get the job done.

[00:18:07] I need just enough of this.

[00:18:08] And I can buy one license or two licenses at a time until I can afford the bigger package because I have a thousand desktops.

[00:18:16] Right.

[00:18:17] So, Carl, I'm going to say I actually disagree with something you just said in there.

[00:18:20] Like you said this idea of like, hey, we have to replenish the money to serve that.

[00:18:24] No, I would argue that that portion of the business is going away.

[00:18:28] Just like we don't go back in time and say, well, you know what?

[00:18:30] We have to replenish.

[00:18:32] New people who come into the industry with one person in a company is going to go away?

[00:18:36] How is that possible?

[00:18:38] No, no, no.

[00:18:38] What I'm getting at is that the role of what they do has changed.

[00:18:43] We do not replenish the money to do things that was done five years ago, that was done 10 years ago.

[00:18:48] And the things that those humans do changes over time.

[00:18:52] I totally agree with the statement of we need to have things for new entrants to do and there needs to be access to that.

[00:18:59] But it is not a consistent same thing over time.

[00:19:02] Right.

[00:19:02] And I think a lot of providers in this space are struggling with is some of the things that they have been doing that they thought were very high value are no longer very high value.

[00:19:12] And, you know, and, you know, and I, you know, the quote unquote fixed computer guy is a lot less valuable than they used to be.

[00:19:21] And I think that that is a hard reality to face.

[00:19:26] That's certainly true.

[00:19:26] I had a cousin who posted something on a forum the other day that was literally like, hey, does anybody know somebody who can fix a keyboard?

[00:19:35] And I'm like, what the hell are you talking about?

[00:19:38] Yes, I can fix a keyboard.

[00:19:40] It's called Amazon.com.

[00:19:41] Right?

[00:19:43] Office Depot, 20 bucks.

[00:19:45] Right, exactly.

[00:19:45] I have fixed your problem.

[00:19:47] No one's going to fix anything.

[00:19:48] So, yeah.

[00:19:49] So the economy is changing.

[00:19:51] I'm just worried about the 95% that is basically being ignored by all of the magazines, all of the stage time, everything in our industry.

[00:20:04] Right?

[00:20:04] There's, you know, the people with a billion dollars have all of the focus in our industry.

[00:20:10] Well, and I think that that's exactly where my observation attaches, Carl.

[00:20:15] Because I think of this in the spirit of anytime we say platform, there are examples in other segments of the industry that are good corollaries.

[00:20:26] And right now I'm looking at the hyperscalers as the cloud infrastructure platforms.

[00:20:30] And the entire industry is enamored with this concept of commits.

[00:20:34] Like you signed up for AWS and you signed a contract that said you would consume X.

[00:20:40] And at the end of the year, you realized you consumed 20% less than X.

[00:20:45] And so everybody else should go and attach themselves and sell against those things.

[00:20:50] And that's your free budget.

[00:20:52] Yeah.

[00:20:52] Except that any customer who is served by any MSP that has four or fewer employees is not large enough to have signed a commit with AWS or any of the other platforms.

[00:21:06] And therefore, all of that conversation about where to go and who to serve and how to get paid and how to build your business model is ignoring 95% of the industry.

[00:21:16] I get the use case.

[00:21:18] I understand the economics of wanting to go after these larger players.

[00:21:22] But I'm looking at it and saying, okay, every time our industry has tried to deploy a platform to automate the servicing of everybody at that SMB level, we have failed miserably.

[00:21:38] Whether it was on service providing, the geek squad, right, was going to nationalize the truck roll for us in the industry.

[00:21:46] And other players have come along and said, well, we're just going to roll up a franchise model and serve everybody.

[00:21:52] And now that MSP services can be effectively delivered remotely, no one will ever sign on with an individual MSP.

[00:22:00] Everybody will be aggregated as part of the telecom or cable system.

[00:22:06] You notice how none of that's happened?

[00:22:08] Okay.

[00:22:08] I think that that's the problem that we're getting into.

[00:22:11] We're building systems to serve the top 1% where the real business case is you're going to aggregate 100% of the audience.

[00:22:20] Only that's never going to happen.

[00:22:23] Well, 100%, but the aggregation model still works at the bottom.

[00:22:27] And I'll sort of end this with saying, look, there's space at the bottom for an aggregator who can do a larger commit, who then subdivides that out to very small businesses.

[00:22:36] This is like, that's a thing, totally viable, will continue to be viable for a very long time.

[00:22:41] But we could spend another hour on this one.

[00:22:44] So I'm going to throw guys, because this one's been percolating for me, topic three.

[00:22:47] And I was like, I reported it over on the Business of Tech.

[00:22:50] And I knew immediately these were data points I had to ask the two of you about.

[00:22:54] My two current favorite data points are in complete collision with one another.

[00:22:57] And I got to throw them your way.

[00:22:59] The first is some Forrester data in a report by Forrester, notably commissioned by Microsoft.

[00:23:04] Early adopters of M365 Copilot among small and medium-sized businesses are reporting benefits, right?

[00:23:12] And they are justifying the $30 per month user cost.

[00:23:16] The survey did 250 SMBs worldwide and found that organizations using Copilot are seeing a 6% increase in revenue, 20% reduction in employee attrition.

[00:23:27] More than half the respondents expect a 20% cut in operating costs, 75% reported improved employee satisfaction, and over 80% anticipated better customer retention.

[00:23:40] Right?

[00:23:41] So, okay, pile of goodness over in the Forrester bucket.

[00:23:44] Sounds like a win.

[00:23:44] Right.

[00:23:45] However, the information is also reporting that 365 Copilot has received a mixed response from customers due to performance issues and high costs, right?

[00:23:54] Leading a lot of businesses to pause adoption.

[00:23:57] They noted that BlackRock and EY have both invested and are expanding its use.

[00:24:02] But the overall adoption rate remains between 0.1% to 1% of Microsoft 365 users paying for the feature.

[00:24:12] I love this because I trust both sources, right?

[00:24:16] I have every reason to believe that Forrester has done good research, and I have every reason to believe the information is also correct.

[00:24:23] The information from the information.

[00:24:24] So, Jens, I know I have thoughts, but I'm dying to hear your thinking on the collision of these data points.

[00:24:33] I'll tell you one thing.

[00:24:35] The Microsoft, the stuff from information about, you know, the larger market, that is not SMB, right?

[00:24:43] Most users of Office 365 and M365 are not SMB.

[00:24:50] And so, you know, what's happening is a lot of those people, I firmly believe they are using some AI, but they're not telling their bosses about it because it might be prohibited.

[00:25:00] It might be forbidden.

[00:25:02] It might be that there's all kinds of restrictions.

[00:25:04] And they want to take credit that they are the geniuses.

[00:25:07] And so, even if they're doing some of the work with AI, they're not using CoPilot and they're not using it through work.

[00:25:14] And so, it's not being adopted.

[00:25:16] That's one other thing to consider.

[00:25:17] Okay, that's solid theory.

[00:25:18] See, and I will observe in all of our engagements working with people, AI is a ramp.

[00:25:25] It is being used.

[00:25:27] It is deployed and it is productive.

[00:25:30] I've been very happy to see how we can be creative, right?

[00:25:34] It writes code, surprisingly well, like legit, surprisingly well.

[00:25:39] It does write meeting notes and it does summarize all of the agenda items and information.

[00:25:46] And it gives you a good editing partner for the content that you are creating, right?

[00:25:51] We all benefit from having an editor and AI does a ruthless job of saying, I don't care who you are.

[00:25:58] I'm going to chop 40% of the words out of the stuff that you just wrote.

[00:26:01] It's very good in those areas.

[00:26:04] But it's almost universally being done as skunkworks.

[00:26:08] Everything that I am seeing and everything that we are observing through vendor implementations as technology companies go,

[00:26:16] everybody's creating use cases.

[00:26:18] Everybody is building out the adoption curve.

[00:26:22] Everybody is saying, here's what you would want to do when we get paid on consumption.

[00:26:26] So let me teach you a thousand ways to use this stuff.

[00:26:30] And all of the formal adoption things are, well, that's difficult and that's bureaucratic and that seems like it's extra and it's really not worth it.

[00:26:39] But over here, my own personal subscription to insert your favorite AI into the blank here, right?

[00:26:47] Everybody's got their own personal tool that they like to use.

[00:26:51] They're paying the 20 bucks a month on their own and using it to improve their own personal productivity and therefore work-life balance and quality of life experience.

[00:27:04] And that, I think, is a use case nobody's paying attention to.

[00:27:08] It's business application, personal funding for professional benefit.

[00:27:16] But does all of that roll up to a use case that will benefit the bottom line of the organization?

[00:27:23] I don't know.

[00:27:24] That's not the way that it's actually being adopted.

[00:27:27] It's being adopted by you and me much the same way that social media was back in its early days.

[00:27:33] I remember the days when Twitter first came out and companies were like, only these four job titles in our organization are authorized to post on Twitter anything about us.

[00:27:43] And we have to go through the professional PR department to vet your message before you tweet.

[00:27:50] And it was like, that's cute.

[00:27:51] There's a billion users and we're all doing it anyway.

[00:27:54] I think we're right there with AI.

[00:27:56] But the personal funding for my job, that was a surprise to me.

[00:28:03] Well, interesting.

[00:28:04] So I'm going to smile and go through those two things.

[00:28:06] First off, social media, that worked out really well.

[00:28:08] Let's just, let's say if we're going to parallel with that, that worked out really well.

[00:28:11] But the second thing is, Ryan, I will actually tell you, like, I'm starting to report on the trend a little bit.

[00:28:16] And it's called shadow AI, is that it is very much following the pattern of shadow IT and that we're seeing a lot of shadow AI.

[00:28:24] Now, of course, this could be potentially incredibly scary for organizations because your data is now just going out into consumer products, like just flowing out.

[00:28:37] You might have copyright violations that the corporations aren't aware of.

[00:28:41] Well, copyright violations on top of copyright violations.

[00:28:44] Inside every single powerpoint.

[00:28:46] You might be using models, right?

[00:28:48] You might be using models that have copyright violations within them, which you are then also committing more copyright violations on top of that.

[00:28:55] Like this is an – so it's interesting to me.

[00:28:58] But I will also tell you I'm starting to see, you know, actual people using these.

[00:29:03] Anyways, I'm going to say this.

[00:29:05] This interview is coming out over the next couple of weeks over on Business of Tech.

[00:29:09] I literally spoke with a vendor when we were talking about other things who then, when I was – you know, I was asking about their AI, who literally brought up Copilot on their own.

[00:29:17] Who said, oh, yeah, we're extensively using Copilot across our organization.

[00:29:20] And so I pressed.

[00:29:21] And it's like not only did they get – we're able to explain exactly all the scenarios they're using in it.

[00:29:25] They are also paying for it.

[00:29:26] They're a subscriber within M365.

[00:29:30] So it's interesting.

[00:29:31] So I want to make sure that I'm also looking at this saying there's a lot going on that we're not necessarily getting great insight into, you know, and whether or not – you know, they may well be within that 1% of Microsoft 365 users.

[00:29:44] That's still a lot of users, right?

[00:29:46] That's still – that's a pretty good platform.

[00:29:47] So let's not be – you know.

[00:29:49] And that means there's a lot of upside here.

[00:29:51] There's a lot of potential.

[00:29:52] Again, I always keep saying, and I like complicated problems because there are opportunities for us to work with customers and give them guidance.

[00:29:59] That's the value.

[00:30:00] But I wanted to bounce this one off of you guys because there's an interesting –

[00:30:04] Do you guys want to – do you want to know what I think is the silliest slash best side effect of all of this?

[00:30:11] Everybody lives in a world of PowerPoint decks.

[00:30:13] There's way too many PowerPoint decks out there, internally, externally.

[00:30:16] And they're always like eight-point type and 40,000 words on a slide.

[00:30:21] They're ugly.

[00:30:22] They're horrible.

[00:30:22] They're wrong.

[00:30:23] AI is getting good at designing PowerPoint slides.

[00:30:26] And in the last six months, I have noticed a distinct trend towards less ugly PowerPoint presentations literally everywhere.

[00:30:36] In everything, whether it's main stage or the boardroom or whatever, we're getting less ugly PowerPoint slides.

[00:30:43] So that's nice.

[00:30:45] Final comment on this.

[00:30:46] I am finding myself using AI because it's built into and has some nice features inside of Adobe Photoshop.

[00:30:53] Right?

[00:30:54] It has some nice features here and there, and I find myself – the only one that I actually pay for is Gemini, which is not, by the way, co-pilot.

[00:31:02] And the second question that I always ask Gemini, and it hasn't learned that this is going to be the second question, is always, do you have citations for that?

[00:31:13] To which it always answers, hey, that's a really good idea.

[00:31:16] You should always ask for documentation of anything, any information I give you.

[00:31:22] But it doesn't offer it up front.

[00:31:24] I can do a whole presentation on my own use of AI, and actually just recently did.

[00:31:31] Yes, and you are the poster child for AI.

[00:31:36] You are literally using it the smartest way of anybody that I know and the most thoroughly of anybody that I personally know.

[00:31:42] So there you go.

[00:31:44] Power user.

[00:31:45] Cool.

[00:31:45] All right.

[00:31:46] With that very happy note, we bring it in to episode 210 of the Killing It Podcast.

[00:31:55] Thanks for tuning in to the Killing It Podcast.

[00:31:58] Please share with your friends and tell everyone to subscribe on iTunes and all the podcast places.

[00:32:06] Join us next week and help us keep killing it in the technology business.